Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Ecosystem Vulnerability to Climate Change 1
1
A Preliminary Assessment of Ecosystem Vulnerability to Climate Change in Panama
Laura Tremblay-Boyer and Eric Ross Anderson
ENVR 451 – Research in Panama
McGill University and the Smithsonian Tropical Research Institution
CATHALAC
Tel: (507) 317-3200
Fax: (507) 317-3299
E-mail: [email protected]
Address: Ciudad del Saber, Edificio 801, Clayton, Panama
Ecosystem Vulnerability to Climate Change 2
2
Table of Contents
II. Host Institution.....................................................................................................................7
III. Study Site ............................................................................................................................8
IV. Methods ...............................................................................................................................8
A. Calculating the index’s components: four component of ecosystem vulnerability to
climate change .......................................................................................................................8
1) Vulnerability to sea level rise – EVCC1........................................................................9
2) Ecosystem geometry (edge effect and irregularity) – EVCC2 ......................................11
3) Climatic “space” of ecosystems – EVCC3 ..................................................................14
4) Species sensitivity – EVCC4 .......................................................................................18
B. Constructing the index: combining four variables into one.......................................21
C. Displaying the index and associated products:...........................................................21
D. Methods for applying index to current situations: .....................................................22
1) Degree of Human Intervention ..................................................................................22
2) Protected areas..........................................................................................................22
3) Species richness and endemism.................................................................................23
V. Results and Discussion .......................................................................................................23
A. Four components of ecosystem vulnerability to climate change ...................................24
1) Vulnerability to sea level rise – EVCC1......................................................................24
2) Ecosystem geometry (edge effect and irregularity) – EVCC2 ......................................28
3) Climatic ‘space’ of ecosystems – EVCC3...................................................................32
4) Species sensitivity – EVCC4 .......................................................................................37
B. Overall EVCC index .......................................................................................................44
C. Applications of the index ................................................................................................49
1) Degree of human intervention........................................................................................49
2) Protected areas .............................................................................................................51
3) Species richness and endemism .....................................................................................54
VI. Conclusions .......................................................................................................................57
X. References ...........................................................................................................................59
Appendix 2. Endemic Species in Panama .................................................................................63
Appendix 3. Special Focus Maps..............................................................................................65
Appendix 4. Chronogram of Activities & Time Spent ..............................................................67
Appendix 5. Notes of Gratitude – Contact Information.............................................................70
Ecosystem Vulnerability to Climate Change 3
3
I. Introduction
Human-induced climate change has shifted status from hypothesis to reality during the
last decade. Global temperature averages are on the rise: of the twelve warmest years since 1850,
eleven of them were between 1995 and 2006, and the (IPCC, 2007) has very high confidence this
is a result of anthropogenic activities. Future scenarios predict global temperature to rise between
1.8 ºC and 4.0 ºC in the next ninety years. Because climate is very much a spatially
heterogeneous variable, the actual effects of global warming at the regional scale are likely to be
much more pronounced. Amongst other things, climate variability, heat waves, droughts and
heavy precipitation events are likely to increase in intensity and frequency, and changes in ocean
currents are expected to have unpredictable consequences (IPCC, 2007).
Climate is one of the key features defining the nature of ecosystems (Hawkins et al.,
2003). Current global warming has already been shown to impact the Earth’s biota at all scales
of organization from species to ecosystems, and the rate of change is likely to accelerate in the
future (Walther et al., 2002), especially if the effects of climate change interact with other drivers
like land-use use change and biotic exchange (Sala et al., 2000). Consequences of global
warming on biotic communities include shifts in species distribution and phenology (Parmesan,
2006; Thomas et al., 2004), increased risk of pathogen outbreaks (Pounds et al., 2006) and
disturbance of predator-prey cycles (Frederiksen et al., 2006). Complex effects resulting from
species interactions have also been documented and will undoubtedly be a common occurrence
given the high number of interacting species forming ecosystems (Ducklow et al., 2007; Walther
et al., 2002). Therefore, as the perceived effects of climate change are very likely to increase in
the future (IPCC, 2007), there is a very tangible risk that there will be important repercussions on
the identity and organization of biological systems.
Ecosystem Vulnerability to Climate Change 4
4
Panama is a neotropical country that is host to a great number of species. Like most
developing countries it suffers from various social and economic issues, and these in turn are
reflected in the state of the environment. For instance, five hundred and seventy-five species are
listed on the 2006 IUCN’s Red List are from Panama (IUCN, 2006). It is thus very likely that
climate change will interact with socio-economic pressures and increase the strain already
present on Panamanian ecosystems. As a signatory to the Convention on Biological Diversity
(CBD), Panama has agreed to take the necessary steps to conserve and protect national
biodiversity in a sustainable manner (Emilio Sempris, pers. comm.). Given its high percentage of
remaining forest cover – 62.4% of the original primary forest (Institute, 2007) – and the large
area of the country covered by protected areas, it is a conservation opportunity with great
potential. For policy-making purposes, it is essential to identify the ecosystems that are most
likely to be affected by climate change in Panama as well as understand how this will impact the
rate of biodiversity loss.
The purpose of our internship is to evaluate the vulnerability of Panamanian ecosystems
to climate change. One of the main challenges is to define what exactly is meant by
“vulnerability” and what its proper measurement is. The IPCC defines vulnerability in terms of
climate change as, “the degree to which a system is susceptible to, or unable to cope with,
adverse effects of climate change, including climate variability and extremes” (IPCC, 2001).
This definition introduces yet another fuzzy term, “susceptibility,” which in turn refers to
exposure, sensitivity and adaptive capacity (Metzger et al., 2005). In this study we apply this
concept of vulnerability to ecosystems, in that vulnerable ecosystems are more likely to suffer
biodiversity loss from climate change. We realize this is still a rather loose definition, and the
Ecosystem Vulnerability to Climate Change 5
5
general lack of understanding of the consequences of climate change for ecosystems makes it
difficult to define straightforward quantifiable responses.
As a preliminary step to assess the vulnerability of biological systems to climate change,
we reviewed methodologies used in both the scientific and institutional literature. We found that
most studies focus on the effects of climate change on biodiversity at the species-level scale. For
instance, there are numerous studies that have modeled predicted shifts in species’ distribution to
infer extinction risk of individual species (e.g., (Thomas et al., 2004; Thuiller et al., 2005b).
However, the relevance of this in terms of species’ ecology is being debated (Hampe, 2004),
especially since the results are strongly influenced by assumptions about species dispersal
abilities (Ibanez et al., 2006; Pearson, 2006). Other methods encountered include field
experiments, game-theory population models, expert opinion and outcome-driven modelling and
scenarios (Sutherland, 2006).
It is one thing to predict the effects of climate change on a species but quite another to try
to predict the consequences on ecosystems, which are made of countless interacting biological
and physical factors. While the relationship between species and ecosystem functions is far from
being understood (Peterson et al., 1998), there is a general consensus that biodiversity loss
decreases the resilience of ecosystems to biotic and abiotic disturbances (Hooper et al., 2005). In
a policy-making framework, it is certainly more useful to be able to assess vulnerability at the
ecosystem scale; however, due to its complexity and inherent uncertainties, this is a task that few
have attempted. For example, (Scholze et al., 2006) have looked at the effects of global warming
on key ecosystem processes (change in forest cover/carbon storage, wildfire frequency, and
freshwater availability) for the world ecosystems by using a Dynamic Global Vegetation Model
(DGVM). Most studies focused on the ecosystem scale actually use either some version of a
Ecosystem Vulnerability to Climate Change 6
6
DGVM alone or combine it with one of the methods mentioned above (e.g., (Schroter et al.,
2005; Thuiller et al., 2006). The short timeframe, the potential amplitude of the task and our
level of expertise prevent us from using this type of methodology. We thus decided to perform a
preliminary vulnerability assessment where we would analyze general characteristics of
ecosystems that are likely to increase their sensitivity to climate change.
The main conservation objective underlining this study is the preservation of all of
Panama ecosystem types in the future, both to minimize biodiversity loss and to maintain
ecosystem services to Panamanians. Using scientific literature, we identified four key
components that could increase ecosystem vulnerability to climate change: (1) sea level rise, (2)
ecosystem geometry (area and shape), (3) climatic “space” and (4) species sensitivity (see the
next section for a detailed description). These four variables will be created and combined using
Geographical Information Systems (GIS) in order to construct an index of ecosystem
vulnerability to climate change (EVCC) for the ecosystems of Panama. Because we are well
aware of the uncertainties associated with a number of ecosystem responses to climate change,
we will not produce a definitive measure but rather rank each ecosystem patch according to its
vulnerability, from our calculations, to climate change under each of the four components. The
EVCC will be a weighted sum of the ranks for each component. It can thus be considered more
of a relative measure: it does not tell how much vulnerability an ecosystem has per se but how
much vulnerability, compared to other ecosystems, it has. The index will be presented as a series
of maps. Additionally, we will look at the relationship between the EVCC, the degree of human
intervention in ecosystems, the current network of protected areas, the distribution of endemic
species, and overall species richness, as these are relevant to Panama’s involvement in the CBD.
While we realize that there are a lot of uncertainties associated with this index, both with the data
Ecosystem Vulnerability to Climate Change 7
7
used and the theory behind it, we feel it will be useful as a preliminary assessment for policy-
making, which will strive to meet Panama’s CBD objectives and improve the robustness of the
protected areas network. This project also has potential to serve as a foundation for continued or
more sophisticated studies on ecosystems and climate change in Panama.
II. Host Institution
The Water Center for the Humid Tropics of Latin America and the Caribbean
(CATHALAC) located in the City of Knowledge, Republic of Panama, is the headquarters of the
Regional Visualization and Monitoring System. CATHALAC operates SERVIR which has a
test bed and rapid prototyping facility managed by the NASA Marshall Space Flight Center at
the National Space Science and Technology Center in Huntsville, Alabama. SERVIR addresses
the nine societal benefit areas of the Global Earth Observation System of Systems (GEOSS):
disasters, ecosystems, biodiversity, weather, water, climate, oceans, health, agriculture, and
energy. It provides tools and technology to monitor resources and environmental conditions by
utilizing satellite imagery and other data sources. Results are presented to various levels of
regional governments and are also made freely available to the general public.
This resource monitoring involves daily and weekly generation of data on current
conditions such as weather, climate, marine environments, natural disasters, and ecological
productivity. Most of the results are used for immediate decision-support; therefore, there has
not been a concentrated effort on performing in-depth analyses of trends, which is what this
particular study will attempt to do.
Ecosystem Vulnerability to Climate Change 8
8
The supervisors for this project are Emil Cherrington, the Geographic Information
Systems and Natural Resource Specialist for SERVIR, and Roxana Segundo, the Development
and International Cooperation Officer for CATHALAC.
III. Study Site
SERVIR shares and analyzes satellite and other types of data for all of Mesoamerica, but
this study focuses on Panama and its biodiversity at the ecosystem level. See the Introduction for
a more specific description of the country in the context of our project.
IV. Methods
In compliance with McGill University’s research ethics statement (2007), appropriate
measures were taken to maintain the integrity of this project. No certification was necessary for
research subjects since all of the work was done at CATHALAC in the computer laboratory;
however, credit is given to those organizations and people who provided our data. Results are
not definitive since the vulnerability assessments are based on climate change scenarios and
projections and because this is only a preliminary study that explores how to evaluate
vulnerability.
A. Calculating the index’s components: four component of ecosystem vulnerability to
climate change
Each of the four components was compared to or is related to a map of ecosystems in
Panama, which had been classified according to vegetation and level of human intervention.
There were thirty-seven ecosystem types and 1303 ecosystem patches (e.g., there are five
“Broadleaf evergreen altimontane rainforest” patches) (Appendix 1). The methodology was to
Ecosystem Vulnerability to Climate Change 9
9
compare all components individually to each of the 1303 ecosystem patches, but generalizations
have been made about the thirty-seven ecosystem types. Since the domain analyses involve
different data types, the overlay methods were different and are described below. The general
processes of analyzing vulnerability in terms of each domain are provided, but the specific
technicalities of GIS have been spared from this paper. Indices were created for each domain,
according to its number (EVCC1,2,3,4)—each EVCCx receiving a value 1-10 (or 0-9 in the case of
EVCC1). They were later combined to create an overall vulnerability index—EVCC.
1) Vulnerability to sea level rise – EVCC1
Land loss due to sea level rise is a direct consequence of climate warming and will likely
impact a number of coastal ecosystems (e.g. mangroves). A digital elevation model (DEM)
provided elevation values at a ninety-meter scale. Since the IPCC (2001) “business as usual”
scenario predicts a 0.6 meter rise in sea level by 2099, all areas from zero to one meter in
elevation were identified. To avoid impractical results, only areas within one kilometer of
coastline were included, which will be called the “coastal buffer zone.” All areas of zero and
one meter elevations are called “red-zones.” The ecosystem patches were then assigned to the
coastal buffer zone, giving each “coastal patch” an identity (Figure 1). The following calculation
was used to determine EVCC1′:
Where cp
rc
A
AR = ,
ep
cp
A
AREVCC ='1 , (Equation 1)
where R is the area of red-zones in all coastal patches of an ecosystem patch (Arc) divided
by the area of all the coastal patches in an ecosystem patch (Acp), and Aep is the area of an
ecosystem patch (Figure 1). It is obvious that Acp is cancelled out in this equation, but it is
retained because of its importance in a further analysis.
Ecosystem Vulnerability to Climate Change 10
10
Figure 1 This example of an ecosystem patch along the Gulf of Chiriqui contains over thirty coastal patches (some
too small to see at this scale). Other ecosystem patches (in different colors) break up patch #832. The tiny green dots
(90m resolution) are marked as red-zones.
EVCC1′ is not the vulnerability index for sea level rise but rather the density of red-zones
in all coastal patches multiplied by the density of coastal patch area in ecosystem patch area.
This means that EVCC1′ is a normalized value of how much red-zone there is in an ecosystem
patch, given only the areas within one kilometer of the coast. In order to determine EVCC1 the
EVCC1′ values were separated using the Geometric Interval quantitative classification scheme in
ArcMap. This scheme was created especially for continuous data and since EVCC1′ is a type of
density, this is a pertinent way to assign vulnerability values to each of the 1303 ecosystem
patches. Values zero to nine were assigned according to these intervals. The rationale behind
not using one to ten is that there are ecosystem patches that should be classified with “zero
vulnerability” to sea level rise because of their geographical location. For the other three
components, ranks of one to ten were used.
Ecosystem Vulnerability to Climate Change 11
11
2) Ecosystem geometry (edge effect and irregularity) – EVCC2
Some ecosystems are likely to be more vulnerable to climatic disturbances brought about
by global warming because of their area and/or shape alone. Larger areas are less susceptible to
ramifications of stochastic perturbations (Janzen, 1983). Also, the shape, more specifically the
ratio of the perimeter to the area of the core, is important because of the edge effect and because
ecotones—transition zones with environmental gradients—are more sensitive to external factors
(Murcia, 1995; Saunders et al., 1991; Shafer, 1999). A number of tools have been developed in
the literature on optimal natural reserve shape, and they have been used to estimate this
component of ecosystem vulnerability to climate change. Two measurements for ecosystem
patch geometry were used. The first was an edge to core ratio calculation, which compares areas
(EVCC2a). The second was an attempt to evaluate of the irregularity of the ecosystem patch
(EVCC2b). A technique similar to EVCC1 was used to create more “buffer zones,” but this time
buffers were created for each ecosystem patch—not just the coastline. Measured inwards from
the border of a patch, a 500 meter “edge” was created for each patch (Figure 2). To calculate
EVCC2a′ a simple density equation was used:
e
ca
A
AEVCC ='2 , (Equation 2)
where Ae is the area of the edge in an ecosystem patch and Ac is the area of the core (even
though the inverse of this equation would obviously give the edge to core ratio, but a frequent
problem of Ac being equal to zero occurred because many patches were simply too small to have
any core). This gave a high value for those ecosystem patches that had a relatively high amount
of core compared to edge. This would mean that a higher EVCC2a′ indicates a lower
vulnerability. To create EVCC2a the values of EVCC2a′ were separated into five groups using the
Ecosystem Vulnerability to Climate Change 12
12
Natural Breaks (Jenks) classification scheme in ArcMap (low EVCC2a is high vulnerability).
Because the Geometric Interval did not distinguish relative differences, the Natural Breaks was
able to break the EVCC2a′ values according to where there were natural boundaries or separations
in the distribution.
Figure 2 These ecosystem patches north of Yaviza, Darien display the 500m-wide edge areas. Comparing edge area
to core area, one can see that #757 has a much higher core:edge ratio than #763 does.
Irregularity was evaluated by measuring the area and perimeter of an ecosystem patch,
converting it into a perfect circle (while retaining the area), and measuring the perimeter of this
circle (Figure 3). The theory behind this method is that a circle has the least perimeter per area
of any shape, which means there are fewer edges that are susceptible to change. After measuring
the area and perimeter of the ecosystem patch and its perfect circle, EVCC2b′ was calculated:
Ecosystem Vulnerability to Climate Change 13
13
circle
ep
bP
PEVCC ='2 , (Equation 3)
where Pep is the perimeter of the ecosystem patch and Pcircle is the perimeter of its perfect
circle. For the same reasons as EVCC2a′ the vulnerability index for EVCC2b was determined by
the Natural Breaks (Jenks) scheme into five intervals.
Figure 3 (not to scale) Conceptual method for obtaining EVCC2b. This shows that the ecosystem patch’s perimeter
is over six times as long as a perfect circle with the same area.
EVCC2 was simply obtained by adding the two methods together:
ba EVCCEVCCEVCC 222 += , (Equation 4 )
which yielded vulnerability values from one to ten.
Ecosystem Vulnerability to Climate Change 14
14
3) Climatic “space” of ecosystems – EVCC3
Ecosystems have evolved to fit specific temperatures and precipitation regimes during the
year. Moreover, some ecosystems exist in regions where there are naturally large variations in
climate within and between years. Each ecosystem can thus be said to be adapted to fit a climatic
“space” that accounts both for average climate (for e.g. average temperature and precipitation)
and climatic variability. Climate change will affect the climatic conditions under which specific
ecosystems occur, and vulnerability should increase the further ecosystems are taken away from
the conditions they have been historically occuring in. Moreover, ecosystems with higher annual
variations in climate should be more resilient to climate change. The third component of the
EVCC, the climatic “space” of ecosystems, uses historical climate data and predicted future
climate to estimate which ecosystem patches, given the predicted change in climate where they
occur and the variation of climate they have been historically exposed to, should be more
vulnerable by climate change. We use temperature and precipitation data in our evaluation of
EVCC3, since they have both been shown to be important determinants of vegetation type
(Hawkins et al., 2003).
The historical climate data was obtained from WORLDCLIM, an electronic database that
provides high-resolution interpolated climatic data (1 km2) for the world, averaged from 1960 to
1990 (Hijmans et al. 2005). The data includes both standard variables like mean temperature and
precipitation but also a set of derived bioclimatic variables, such as the precipitation of the driest
quarter, that are more relevant to biological systems. The historical data obtained from
WORLDCLIM were temperature and precipitation for the months of July through September as
well as a number of bioclimatic variables described below.
Ecosystem Vulnerability to Climate Change 15
15
The predicted future climate data for Panama was obtained from CATHALAC’s climate change
model (CCCM). The model was run at a scale of 36 km2 from 2005 until 2099. The data used in
EVCC3 was monthly temperature and precipitation average for july through september of the
years 2025, 2050 and 2099. The analysis was focused on those months because of data
availability but also because focusing on one season of a region with a lot of inter-annual
variation in climate allows to capture more variation. In the case where data from other parts of
the year become available and a third party would want to run the EVCC with them, we would
suggest not averaging the data together but doing separate analyses for biologically relevant
periods of the year. Of course, the magnitude of change in one period is likely to be correlated
with that in another and this would have to be accounted for. Additionally, as for now CCM has
only been run with one carbon emission scenario “business-as-usual”. It would have been
preferable to evaluate and compare the results of the EVCC according to a number of scenarios,
but this caveat is balanced by the fact that CATHALAC’s climate change model is the one with
the highest resolution for that region. Low-resolution is often invoked as one of the main
problems in studies of ecological response to climate change. CATHALAC’s climate change
model will be run with more scenarios in the future and other parties will have the possibility to
use the framework we have set up for the “business-as-usual” to evalutate EVCC under different
scenarios.
The analysis itself was conducted at a resolution of 1 km2, using the WORLDCLIM data as a
template. As a result, each 36 km2 pixel of the CCM were divided in 36 1 km
2 pixels which were
assigned the climate values of the CCM pixel they were derived from. While this induced a bias
in the EVCC3 in terms of spatial auto-correlation of the up-scaled pixel, it allowed to include a
greater number of ecosystem patches in the analysis and still provided information about the
Ecosystem Vulnerability to Climate Change 16
16
magnitude of variation in climate for each 1 km2 pixel since the WORLDCLIM data occurs at
that resolution. A map of the excluded ecosystem patches is included in the “Results and
discussion” section (Figure 7).
The Gower Metric was used to evaluate predicted change in climate in terms of temperature and
precipitation for each 1 km2 pixel and the values obtained at the pixel level were averaged at the
ecosystem patch level. The Gower Metric is a measure of dissimilarity between environmental
coordinates that is often used in ecology, for example in the DOMAIN model of species
distributions. The Gower Metric increases as difference in climate increases and as historical
range in climate decreases. In other words, a pixel is deemed more vulnerable if it is predicted to
have a higher change in temperature and/or precipitation with the ecosystem patch were it occurs
having been exposed on average over 1960 – 1990 to a smaller intra-annual range of temperature
and precipitation.
The following steps were followed to build EVCC3 (see figure 4):
a. Climatic data was compilated for each 1 km2 pixel: historical data – mean monthly tempeture
(°C) and precipitation (mm) for the months of July, August and September, mean temperature of
the warmest and coldest quarters (three months), mean precipitation of the wettest and driest
quarters; predicted data – mean monthly temperature and precipitation for July, August and
September 2025, 2050 and 2099.
b. For both historical and predicted climate, mean monthly temperature and precipitation were
averaged over the months of July through September.
c. For each ecosystem patch, the minimum mean temperature of the coldest quarter (Tmin), the
maximum mean temperature of the warmest quarter (Tmax), the minimum mean precipitation for
the driest quarter (Pmin) and the maximum mean precipitation of the wettest quarter were
Ecosystem Vulnerability to Climate Change 17
17
identified (Pmax). These values were substracted to calculate the temperature and precipitation
range for each ecosystem patch.
d. The Gower Metric for temperature and precipitation was calculated for each pixel for 2025,
2050 and 2099 by using the historical and future climate data compiled in Step 1 and the
temperature and precipitation ranges of the ecosystem patch where the pixel occurs calculated in
Step 2. An example for temperature is given here:
Gower Metric for cell i = (temphistorical – tempfuture )/ temprange of ecosystem patch containing i
e. A Gower Metric for temperature and precipitation for 2025, 2050 and 2099 was derived for
each ecosystem patch by averaging the Gower Metric within each category (temperature and
precipitation) of each of the pixels that occur in the patch.
f. For each ecosystem patch, an aggregated Gower Metric for temperature and precipitation was
calculated.
GM ecosystem patch = 4* GM2025 + 2* GM2050 + GM2099 (Equation 5)
This pattern of weighting was chosen because, given all else is equal, an ecosystem where the
rate of climate change is higher should be more vulnerable. Also, uncertainty in the accuracy of
the predicted climate data increases with time.
g. The aggreagated Gower Metrics for temperature and precipitation were compared amongst all
ecosystem patches and ecosystem patches were ranked from 1 to 10 using a … classification
scheme. A score of 10 was assigned to the highest Gower Metrics.
Ecosystem Vulnerability to Climate Change 18
18
h. EVCC3 for each ecosystem patch was obtained by summing the ranks of that ecosystem patch
for the aggregated Gower Metric of temperature and precipitation. Under EVCC3 thus, an
ecosystem patch with a score of 20 was the most vulnerable compared to other ecosystem
patches in terms of both temperature and precipitation. Note that in the way EVCC3 is built,
temperature and precipitation were given equal weight in terms of their effect on ecosystem
vulnerability. This weighting is flexible and could be adjusted in future runs of the EVCC if
deemed relevant.
4) Species sensitivity – EVCC4
A large number of studies have modelled shifts in species distribution in response to
climate change and have gradually uncovered many basic characteristics that could make species
more vulnerable. These include for example geographic extent and basic niche properties
(Thuiller et al., 2005a), thermal range (Jiguet et al., 2006) and life form (for plants) (Broenniman
et al., 2006). Here a geographic extent of representative groups of vertebrates as defined by
NatureServe’s Infonatura (2007) was used for mammals, birds and amphibians to estimate
species’ sensitivity to climate change in each ecosystem. (Thuiller et al., 2005a)s’ study (2005a)
was consulted as a theoretical foundation since they found that species with smaller ranges were
less likely to have suitable habitat available for them in the future. Thus, in this fourth domain
ecosystem types that have a higher number of species with small ranges will be deemed more
vulnerable to climate change.
Estimated ranges for mammals, amphibians and birds of Panama were obtained from Infonatura,
an electronic database that provides conservation and education resources on vertebrates of Latin
America and the Caribbean (InfoNatura, 2007). The resulting coverage of Panamanian vertebrate
Ecosystem Vulnerability to Climate Change 19
19
diversity was high: geographic information on ranges was available for 245 out of the 246
mammals, 845 of the 929 birds (including migratory birds), 192 out of the 196 amphibians.
Steps followed to build EVCC4:
A. Separate analysis of species sensitivity were performed for each group (mammals, birds,
amphibians) because of their inherrent different in range sizes (birds tend to have much bigger
ranges then amphibians) (see figure 4).
1. Total distribution area of species were calculated, including where the species occurs
outside of Panama, and assigned a rank from 1 to ‘x’ according to a quantile
classification scheme where a score of x was given to the 10% smaller, and thus most
sensitive, ranges. ‘x’ was selected according to the number of species in the group to
allow a finer definition of range categories, with ‘x’ being the highest in birds and the
smallest in amphibians.
2. From the species distribution we derived for each ecosystem patch which species
were theoretically occuring there. The average of the ranks calculated in step 1 of all
species of each group that occur within each ecosystem patch was calculated. In other
words, the average of the ranks for all the birds occuring in a given ecosystem patch
was calculated, the average of the ranks for all the amphibians, etc.
3. The species density of each ecosystem patch was calculated by dividing the total
number of species from each group that occur in the patch by the area-rank of the
patch from 1 to 10, as determined by ranking all ecosystem patches area with a
natural breaks (Jenks) scheme.
Ecosystem Vulnerability to Climate Change 20
20
4. For each species group, the average rank divided by the species density for each
ecosystem patch were ranked on a scale of 10 using a geometric classification
scheme. This second ranking was necessary so that the different number of classes
assigned to each group in the initial ranking (see step 1) would not bias the resulting
average species sensitivity per ecosystem patch. The ranks from all three groups were
then added for each ecosystem patch.
B. EVCC 4 was calculated by summing for each ecosystem patch the ranks from all three groups
obtained in the last step and ranking (one last time) the aggregated average ranks according to a
quantile classification scheme. A rank of 5 was awarded to the ecosystem patches that had, on
average, the mammals, birds and amphibians with the smallest ranges for their respective groups.
Figure 4: An outline of the main steps leading to the final calculation of EVCC4
1. Distributions are ranked according to
their area relative to that of other species
of their group
2. For each ecosystem patch and for each of the
three groups, the average of the ranks of the
species occuring in the patch is calculated
3. The resulting average rank for each group is
divided by species density and ranked against
those of the other ecosystem patches.
Distributions of all mammals, birds, and amphibians species
found in Panama
Ecosystem Vulnerability to Climate Change 21
21
B. Constructing the index: combining four variables into one
For each ecosystem type appropriate weights were assigned to each of the four
components defined above. This is important because some factors are not as pertinent to climate
change or cannot be as reliable in either the raw data or the analyses (Table 1).
Table 1 How to rank EVCC components: Numbers in italics represent the more influential
category for each domain, which is the deciding factor in the case of ties.
Domain
Pertinence to
Climate
Change
Reliability of
Data
Overall
Weight Scale
EVCC1 (sea level rise) 3 4 3 0 – 9
EVCC2 (patch geometry) 1 2 2 1 – 15
EVCC3 (climatic ”space”) 4 3 4 1 – 20
EVCC4 (species
sensitivity) 2 1 1 1 – 5
Considering that climatic ‘space’ is at the heart of this study, it has received the highest
weight. Sea level rise follows in second, and ecosystem geometry is the third most important.
Species sensitivity has been given the lowest weight because of the uncertainties in the theory of
species range as a driver of ecosystem patch vulnerability. There is also the problem of relying
on this species presence/absence data, because there remain many parts of Panama that are quite
unexplored. Given these ranks, the EVCC value for all of the ecosystem patches (except some
due to resolution issues, see EVCC3 results) is calculated:
4321 EVCCEVCCEVCCEVCCEVCCtotal +++= . (Equation 6)
C. Displaying the index and associated products:
The EVCC index has been displayed on a simple map of Panama divided into ecosystem
patches. Darker colors generally indicate higher vulnerability; refer to the legend accompanying
each map. Similar maps are also available for each domain of ecosystem patch vulnerability.
Ecosystem Vulnerability to Climate Change 22
22
D. Methods for applying index to current situations:
For discussion and application purposes, the EVCC index was overlain on other types of
data. Doing this allows viewers to pinpoint areas that are especially vulnerable from the lack of
protection and/or the intrinsic biotic vulnerability (due to the endemism), and this will be a useful
decision-support tool in regards to the Panama’s CBD objectives and its focus on protected areas
and endemic species. The degree of human intervention, protected areas, endemism, and species
richness are four applications.
1) Degree of human intervention
Degrees of intervention were assigned to each ecosystem patch (values from zero to nine), which
reflect their UNESCO description. Table 2 guided the method in valuation:
Table 2 Guide to Assigning Intervention Values
Ecosystem Patch Description Intervention Value
Populated places 9
Shrimp farming 8
Agroforestry 7
Productive system: <10% natural or spontaneous vegetation 6
Productive system: 10-50% natural or spontaneous vegetation 5
Natural system: moderately high intervention (lowlands) 4
Natural system: moderately high intervention (mountains) 3
Natural system: medium intervention 2
Natural system: low intervention 1
All others 0
Overall EVCC was compared to degree of intervention to see which domain is the cause for the
most concern.
2) Protected areas
The overall EVCC scores for each ecosystem patch were again compared to a map of the
anthropological reserves, biological corridor, biological reserve, forest reserves, hydrological
protection zones, national marine parks, national parks, national wildlife refuge, private reserve,
Ecosystem Vulnerability to Climate Change 23
23
protected forests, protective zone, recreation area, and wildlife refuges of Panama provided by
CATHALAC.
3) Species richness and endemism
The InfoNatura (Nature Serve, 2007) database provided the names of all mammal, bird
and amphibian species occuring in Panama as well as those that are endemics. Sixteen
mammals, eight birds, and twenty-nine amphibians are endemic to the country (Appendix 2).
Panama was divided into a grid of 10km x 10km squares and species richness and endemism
were calculated for each.
V. Results and Discussion
The results are presented in a similar framework as the Methods section. Discussions on
each set of results are provided directly after each results section so more congruent ideas can be
deduced. Since the focus of this assessment was done on the ecosystem patch level, results have
been generated for parts A and B that contain EVCCs for 1303 patches, but they are not provided
in this paper for the vast amount of space they require. For a more general analysis, summarized
results are also provided at the ecosystem level, in which thirty-seven ecosystem types have been
described in terms of different vulnerability components. A discussion on the different
impressions people can get from these maps is given in part C, and implications of our analysis
of further applications are addressed. For all results, critiques of the methods are provided in
order to initiate thoughts on how improvements can be made. Also, a large table of minimums,
maximums, averages, and variances, of EVCC1,2,,3,4 and overall EVCC summarized for each
ecosystem type is presented in Table 3.
Ecosystem Vulnerability to Climate Change 24
24
A. Four components of ecosystem vulnerability to climate change
1) Vulnerability to sea level rise – EVCC1
A glance at Table 3 shows average EVCC1 values on a scale of 0-14. All of the
ecosystem types with a rank of zero have no land that is zero to one meter in elevation that is on
the one kilometer-deep coastline. Many of these individual patches are landlocked. A six-way
tie for the most vulnerable ecosystem patches (in the maximum EVCC1 column) goes to patches
of the following ecosystem types: mangroves, semideciduos tropical forests in lowlands (all
three levels of human intervention), small islands, and populated places. Sixty-two patches have
the highest EVCC1 value of 14, fifty-four of them being small islands. Over half of these small
islands are in Bocas del Toro, a province that also contains the other most vulnerable ecosystem
types such as mangroves, various types of forests in lowlands, swampy or marshy forests, and
occasionally flooded rainforests. Bocas is also the location of the only populated place patch
with the highest possible EVCC1 value. As seen in this province, often these ecosystem types are
neighbors to one another, which Map 1 attempts to display. The Colón and Free Trade Zone area
is a mix of similar ecosystem types with high patches of vulnerability. An almost identical mix
of ecosystem types with patches of high vulnerability exists along the Gulf of Chiriquí.
In terms of average EVCC1 the two most vulnerable ecosystem types (broadleaf
evergreen rainforests dominated by palms and in swamps, and coastal vegetation growing on
very new soils) only have seven and five patches, respectively, and account for an extremely
small percentage of total land in Panama. This could be a cause for concern because of the rarity
of this ecosystem type. See Figure 5 for a histogram of these results. After mangroves and
islands, which have already been discussed, come the salt and shrimp production ecosystem
types. Again these patches account for a very small area of the country and are located in the
Ecosystem Vulnerability to Climate Change 25
25
Arco Seco of the Azuero Peninsula. They are still worth keeping in mind for their importance in
the economies and food production for the country (discussed in Applications section).
The method of selecting land of only zero to one meter in elevation involved using a
DEM on the 90m scale. Unfortunately is has a ±15m margin of error and slight imperfections
when overlaying it with the ecosystem map, some results are not completely accurate. Still, the
process identified the many of those tiny patches that would be more affected by this kind of
discrepancy and often labeled them as highly vulnerable. Also, by selectively choosing land that
was one kilometer or closer to the coast, some areas that of zero to one meter high were
disregarded. It could have be that inland valleys go even below sea level (and are not necessarily
vulnerable to sea level rise) or that expansive river deltas were wrongfully erased. The largest
bulk of elevations zero to one meter high that were erased occurred in the Arco Seco of the
Azuero Peninsua (dominating the shrimp and salt production). Whether or not this area should be
classified as more vulnerable, it has already received attention for being ranked highly.
Map 1 of vulnerability to sea level rise well-illustrates the particular coastal regions that
are more vulnerable than inland areas. Since the color scheme was applied to ecosystem patch
and not simply which coasts are susceptible to flooding, there is the possibility of misconception
in this demonstration of vulnerability. Most noticeable is the second largest ecosystem patch
(Productive system with less than 10% natural or spontaneous vegetation) that covers a large
portion of the Azuero Peninsula. Because many of its boundaries are coasts, there is a high
chance that parts of it will be susceptible to sea level rise. This shows up in the fact that it was
given an EVCC1 value of 3. Had this patch been divided into smaller sub-patches, a definite
difference in perception would occur.
Ecosystem Vulnerability to Climate Change 26
26
Figure 5 Distribution of average EVCC1 values for each ecosystem type.
0 1 2 3 4 5 6 7 8 9
2b. BSOT latifoliado submontano - BI
2c. BSOT latifoliado submontano - PI
3a. BSOT latifoliado montano
3b. BSOT latifoliado montano - BI
3c. BSOT latifoliado montano - PI
4a. BSOT latifoliado altimontano
4b. BSOT latifoliado altimontano - Med. I
5. BSOT latifoliado nuboso
6b. BSOT aluvial ocasionalmente inundado - Med. I
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria
14. Plantaciones forestales
17. Vegetación de páramo
18a. Pant. de cyperáceas con abund. acum. de mat. org.
20. Flujo de lava con escasa vegetación
23. Carrizales pant. y form. sim. domin. por Typha
16. Sabanas arbol. de graminoides cortos inund.
1c. BSOT latifoliado de tierras bajas - PI
18b. Pantanos herbáceos salobres
8b. B. semideciduo tropical de tierras bajas - bastante intervenido
8a. B. semideciduo tropical de tierras bajas
10. B. deciduo por la sequia, latifoliado de tierras bajas
7a. BSOT pantanoso - domin. por dicotiledoneas
22. Albina con escasa vegetación
8c. B. semideciduo tropical de tierras bajas - poco intervenido
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %)
6a. BSOT aluvial ocasionalmente inundado
1a. BSOT latifoliado de tierras bajas
1b. BSOT latifoliado de tierras bajas - BI
12. Poblados
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %)
7c. BSOT pantanoso - domin. por Campnosperma
15. Sist. prod. acuático-camaronera y salina
11. Isla menores de 140 ha
9. Manglares
21. Veg. costera de trans. sobre suelos marinos muy rec.
7b. BSOT pantanoso - domin. por palmas
EVCC1 rank per ecosystem type
Ecosystem Vulnerability to Climate Change 27
27
Map 1
Ecosystem Vulnerability to Climate Change 28
28
2) Ecosystem geometry (edge effect and irregularity) – EVCC2
It should first be pointed out that the distribution for the average EVCC2 (on a scale of 1
to 10) is not as widely spread as EVCC1. Despite this, ranges still exist at the extremes—EVCC2
of two to ten. Probably that EVCC2 is the sum of two sub-EVCC2 values with often opposite
values is because the average range is very concentrated from five to eight. Also, a value of one
is not possible because the minimum values for EVCC2a and EVCC2b were both one. Adding
two of the least vulnerable sub-values would mean that ‘2’ is the lowest possible EVCC2.
Gathering maximum values from Table 3, one can see that many of the coastal ecosystem
types are again highly vulnerable. This means that their shapes are of the most irregular
(probably long and thin in this case) and they are relatively small (having a large edge to core
ratio). A look at Map 2 shows that generally, ecosystem patches are smaller along the coastline,
likely because the change in land type can increase more dramatically from coast to inland than
from inland to inland. Agricultural systems also have some patches of very high vulnerability in
terms of geometry, but there is a caveat here. The geometry analysis meant to capture those
ecosystem patches with the most susceptibility to negative edge effects. Even though an
agricultural system could have an irregular shape, the theory behind its vulnerability does not
stand. That is to say, this geometrical analysis is much more pertinent for natural systems,
because human-altered systems are often very buffered from surrounding environmental changes
(later discussed in the Human Intervention section).
Ranking the ecosystem patches by highest EVCC2 (see for where to find complete
results) shows that nine out of ten patches are mangrove forests. It more is important to look at
ecosystem patch here because many ecosystem types have very few (and important to this
geometric analysis—small) patches. This is why the histogram (Figure 6) shows coastal
Ecosystem Vulnerability to Climate Change 29
29
vegetation growing on very new soils and salt marshes with scarce vegetation to have the highest
EVCC2 ranks. They are definitely still important to consider because this type of ecosystem has
already been shown to be vulnerable to sea level rise.
The least vulnerable, which should be in theory the most resistant to surrounding
perturbations, include again many patches of lowland forest but also numerous patches in
mountainous regions and agricultural patches. The fact that lowland type ecosystems have
patches that have been classified as both least and most vulnerable further support the argument
that this geometry analysis is not good at summarizing vulnerability based on ecosystem type. It
is much more meaningful when considered for each patch.
Ecosystem Vulnerability to Climate Change 30
30
Figure 6 Distribution of average EVCC2 values for each ecosystem type.
0 1 2 3 4 5 6 7 8 9
3c. BSOT latifoliado montano - PI
4b. BSOT latifoliado altimontano - Med. I
6b. BSOT aluvial ocasionalmente inundado - Med. I
23. Carrizales pant. y form. sim. domin. por Typha
17. Vegetación de páramo
3a. BSOT latifoliado montano
2b. BSOT latifoliado submontano - BI
2a. BSOT latifoliado submontano
20. Flujo de lava con escasa vegetación
10. B. deciduo por la sequia, latifoliado de tierras bajas
3b. BSOT latifoliado montano - BI
4a. BSOT latifoliado altimontano
5. BSOT latifoliado nuboso
1a. BSOT latifoliado de tierras bajas
7c. BSOT pantanoso - domin. por Campnosperma
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria
8b. B. semideciduo tropical de tierras bajas - bastante intervenido
11. Isla menores de 140 ha
8a. B. semideciduo tropical de tierras bajas
1c. BSOT latifoliado de tierras bajas - PI
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %)
14. Plantaciones forestales
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %)
8c. B. semideciduo tropical de tierras bajas - poco intervenido
6a. BSOT aluvial ocasionalmente inundado
1b. BSOT latifoliado de tierras bajas - BI
18a. Pant. de cyperáceas con abund. acum. de mat. org.
12. Poblados
15. Sist. prod. acuático-camaronera y salina
7a. BSOT pantanoso - domin. por dicotiledoneas
7b. BSOT pantanoso - domin. por palmas
16. Sabanas arbol. de graminoides cortos inund.
18b. Pantanos herbáceos salobres
9. Manglares
22. Albina con escasa vegetación
21. Veg. costera de trans. sobre suelos marinos muy rec.
EVCC2 rank per ecosystem type
Ecosystem Vulnerability to Climate Change 31
31
Map 2
Ecosystem Vulnerability to Climate Change 32
32
3) Climatic ‘space’ of ecosystems – EVCC3
EVCC3 is an indicator of how much climate will change on average for each ecosystem patch,
weighted by how much interannual variation in climate an ecosystem patch has been exposed to
during the last 40 years. Only 770 out of the 1303 ecosystem patches were evaluated because of
the resolution of the climate data and specificities of the type of spatial format (raster) that was
used for the analysis. Figure 7 shows highlighted in light blue the ecosystem patches that were
not included in the analysis. Their total area is less than 0.16% of the total country superficie so
that their exclusion from EVCC3 has minimal consequences on the overall EVCC.
Figure 7: Ecosystem patches (in light blue) that were not evaluated under EVCC3
A number of trends can be extracted by observing the EVCC3 maps (Map 3). First, the most
vulnerable regions of Panama, both in terms of temperature, precipitation and overall EVCC3,
are the Pacific and Carribean side of the west part of the country. The province of Bocas del
Toro shows the highest vulnerability overall. In general, the center of the country receives low
vulnerability rank, including the Peninsula de Azuero but excluding some ecosystem patches in
the Darien and around the Canal that are ranked with intermediate-high vulnerability. Given that
a great proportion of these systems are highly exploited for agriculture, this is a positive result in
economic terms. Isla Coiba has intermediate vulnerability, but its status as an island makes the
micro-climate less predictable. In general, bigger ecosystem patches are less vulnerable, making
Ecosystem Vulnerability to Climate Change 33
33
++
EVCC3: Climate ‘space’ of ecosystems Map 3
Ecosystem Vulnerability to Climate Change 34
34
an important proportion of the country’s superficie low vulnerability. The trend can be observed
at the ecosystem type level as well, as shown in figure 8. Figure 8 also shows that no ecosystem
type has on average very high vulnerability for both precipitation and temperature. Visually
comparing the temperature and precipiration maps shows that vulnerability in terms of
temperature and precipitation are well spatially correlated, except for the Canal area that has
more vulnerability in terms of precipitation and the Darien where vulnerability to temperature
change is much more important than the one to precipitation change. This trend also holds when
averaging to the ecosystem type level, as presented in figure Figure 9 which ranks ecosystem
types according to overall EVCC3 total but details the ranking (out of 10) for precipitation and
temperature. When a difference is observed, it is in general temperature that ranks higher, except
for mountain ecosystems where the opposite occurs. Figure 10 details the number of ecosystem
patches that have the same combination of rank for temperature and precipitation and shows that
the correlation holds only at smaller ranks. In other words, ecosystem patches that are extremely
vulnerable in terms of precipitation are generally not that vulnerable to temperature but
ecosystem patches that have moderate vulnerability often rank similarly in terms of temperature
and precipitation.
Ecosystem Vulnerability to Climate Change 35
35
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
4b. BSOT latifoliado altimontano - Med. I – 0.09%
2c. BSOT latifoliado submontano - PI – 0.06%
3c. BSOT latifoliado montano - PI – 0.07%
4a. BSOT latifoliado altimontano – 0.52%
3b. BSOT latifoliado montano - BI – 0.61%
5. BSOT latifoliado nuboso – 0.29%
2a. BSOT latifoliado submontano – 9.12%
8c. B. semideciduo tropical de tierras bajas - poco intervenido – 0.79%
2b. BSOT latifoliado submontano - BI – 1.14%
22. Albina con escasa vegetación – 0.01%
10. B. deciduo por la sequia, latifoliado de tierras bajas – 0.09%
3a. BSOT latifoliado montano – 2.54%
8a. B. semideciduo tropical de tierras bajas – 9.51%
8b. B. semideciduo tropical de tierras bajas - bastante intervenido – 6.08%
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %) – 23.36%
15. Sist. prod. acuático-camaronera y salina – 0.1%
1b. BSOT latifoliado de tierras bajas - BI – 8.8%
18b. Pantanos herbáceos salobres – 0.03%
1a. BSOT latifoliado de tierras bajas – 14.35%
1c. BSOT latifoliado de tierras bajas - PI – 1.59%
7a. BSOT pantanoso - domin. por dicotiledoneas – 0.06%
16. Sabanas arbol. de graminoides cortos inund. – 0.06%
20. Flujo de lava con escasa vegetación – 0.08%
12. Poblados – 0.28%
23. Carrizales pant. y form. sim. domin. por Typha – 0.25%
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %) – 15.34%
14. Plantaciones forestales – 0.07%
6b. BSOT aluvial ocasionalmente inundado - Med. I – 0.04%
11. Isla menores de 140 ha – 0.11%
9. Manglares – 2.86%
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria – 0.18%
17. Vegetación de páramo – 0.03%
21. Veg. costera de trans. sobre suelos marinos muy rec.– 0.05%
18a. Pant. de cyperáceas con abund. acum. de mat. org.– 0.02%
6a. BSOT aluvial ocasionalmente inundado – 1.04%
7b. BSOT pantanoso - domin. por palmas – 0.05%
7c. BSOT pantanoso - domin. por Campnosperma – 0.32%
EVCC3 rank per ecosystem type
Figure 8: Average EVCC3 per ecosystem type with proportion of Panama occupied by ecosystem
Ecosystem Vulnerability to Climate Change 36
36
0 1 2 3 4 5 6 7 8 9 10
4b. BSOT latifoliado altimontano - Med. I – 0.09%
2c. BSOT latifoliado submontano - PI – 0.06%
3c. BSOT latifoliado montano - PI – 0.07%
4a. BSOT latifoliado altimontano – 0.52%
3b. BSOT latifoliado montano - BI – 0.61%
5. BSOT latifoliado nuboso – 0.29%
2a. BSOT latifoliado submontano – 9.12%
8c. B. semideciduo tropical de tierras bajas - poco intervenido – 0.79%
2b. BSOT latifoliado submontano - BI – 1.14%
22. Albina con escasa vegetación – 0.01%
10. B. deciduo por la sequia, latifoliado de tierras bajas – 0.09%
3a. BSOT latifoliado montano – 2.54%
8a. B. semideciduo tropical de tierras bajas – 9.51%
8b. B. semideciduo tropical de tierras bajas - bastante intervenido – 6.08%
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %) – 23.36%
15. Sist. prod. acuático-camaronera y salina – 0.1%
1b. BSOT latifoliado de tierras bajas - BI – 8.8%
18b. Pantanos herbáceos salobres – 0.03%
1a. BSOT latifoliado de tierras bajas – 14.35%
1c. BSOT latifoliado de tierras bajas - PI – 1.59%
7a. BSOT pantanoso - domin. por dicotiledoneas – 0.06%
16. Sabanas arbol. de graminoides cortos inund. – 0.06%
20. Flujo de lava con escasa vegetación – 0.08%
12. Poblados – 0.28%
23. Carrizales pant. y form. sim. domin. por Typha – 0.25%
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %) – 15.34%
14. Plantaciones forestales – 0.07%
6b. BSOT aluvial ocasionalmente inundado - Med. I – 0.04%
11. Isla menores de 140 ha – 0.11%
9. Manglares – 2.86%
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria – 0.18%
17. Vegetación de páramo – 0.03%
21. Veg. costera de trans. sobre suelos marinos muy rec.– 0.05%
18a. Pant. de cyperáceas con abund. acum. de mat. org.– 0.02%
6a. BSOT aluvial ocasionalmente inundado – 1.04%
7b. BSOT pantanoso - domin. por palmas – 0.05%
7c. BSOT pantanoso - domin. por Campnosperma – 0.32%
Average Rank Precipitation Average Rank Temperature
Figure 9: Average rank for temperature and precipitation per ecosystem type and proportion of Panama
occupied by ecosystem
Ecosystem Vulnerability to Climate Change 37
37
0
1
2
3
4
5
6
7
8
9
10
11
0 1 2 3 4 5 6 7 8 9 10 11
Ecosystem rank Temperature
Ecosystem
rank Precipitation
Figure 10: EVCC3 rank for temperature vs. precipitation, area of bubble represents the number of ecosystem
patches with that combination of rank
The trends gathered from EVCC3 show that the vulnerability to climate change in terms of the
climatic ‘space’ of ecosystems is asymmetrically distributed across the country and that it is
wrong to assume all ecosystems will react similarly. Such analysis is not only useful for
conservation purposes but also to increase the resilience of the agricultural system to global
warming.
4) Species sensitivity – EVCC4
EVCC4 is a measure of how many species sensititive to climate change an ecosystem patch
contains compared to other ecosystem patches, accounting for species density as a function of
species richness and area of the patch. EVCC4 was evaluated for all 1304 ecosystem patches. A
map of EVCC4 is presented in figure x, please refer to Appendix X for more detailed results.
Figure x shows average EVCC4 for each ecosystem type and figure x divides EVCC4 into its 3
components, birds, amphibians and mammals, per ecosystem type.
Ecosystem Vulnerability to Climate Change 38
38
An interesthing trend is that mountains ecosystems appear more vulnerable. The north-west part
of the country is also more vulnerable, although this might be partly biased from the fact that the
biggest ecosystem patch was very species rich and that accounting for species density did not
completely eliminate the area effect. Islands generally got lower scores, this is probably due to
their low species density compared to the mainland. It would be interesthing to look more closely
at endemic species in islands as these are likely to be sensitive to climate change. It is
interesthing to visually compare the map of endemism in Panama (see page 56) with the map of
EVCC4. EVCC4 appears to be correlated with endemism, which would be expected given that
endemic species should have the smallest ranges of species within their groups. Finally, figure x2
tells us that species sensitivity ranks of mammals, amphibians and birds are not necessarily
correlated with themselves or with EVCC4, although mammals and birds are generally more
similar per ecosystem type.
The methodology used to measure EVCC4 appears appropriate. As shown in Table 2, accounting
for species density per ecosystem patch allowed to remove the area effect that occurs when one
only considers the average number of sensitive species. This effect resuts from the fact that a
Ecosystem Vulnerability to Climate Change 39
39
EEVVCCCC44 –– SSppeecciieess sseennssiittiivviittyy
Map 4
Ecosystem Vulnerability to Climate Change 40
40
larger area can contain relatively more species with smaller ranges. It would probably be
necessary to explore how the choice of the classification scheme (quantile and geometric) affects
the resulting EVCC4. It might prove worthwhile to justify a more scientifically grounded choice
of classification, that might be more appropriate to the type of data used and/or biologically
relevant. It is important to keep in mind that EVCC4 can only be as good as the initial species
distributions used to construct it so that care must be taken to have distributions as accurate as
possible to start with.
Due to data availability reasons, EVCC4 only looked at birds, mammals and amphibians. While
this yields interesthing results, uncertainty still remains as to how extinctions affect ecosystem
stability and what the relationship is between ecosystem diversity, stability and function (Hooper
et al., 2005). In other words, it is uncertain that a higher number of species that are sensitive to
climate change will make necessarily make an ecosystem more vulnerable to it. It is also
important to keep in mind that climate change is one of the many threats facing species and thus
EVCC4 cannot be considered in isolation for specific species. For instance, the puma is the
mammal with the widest distribution and as a result was not classified as sensitive. However, it is
extremely vulnerable to other drivers and already has disappeared from a number of ecosystems.
Ideally, EVCC4 would be built on actual distribution maps, not on potential, as what it aims at
ultimately is to estimate the risk of extinction and including species that have already
disappeared from the ecosystem add another bias to the measure. Additionally, a possible way to
make EVCC4 less uncertain in terms of ecosystem vulnerability and more relevant in terms of
the UNESCO’s ecosystem classification would be to look at tree distributions. Focusing on the
Ecosystem Vulnerability to Climate Change 41
41
primary trophic level might also be more ecologically relevant when one is trying to link
sensitivity and risk of extinction to ecosystem stability.
The theory behind EVCC4 , while intuitive, is still new (Thuiller et al., 2005a) and is itself very
uncertain. It will be important to keep up-to-date with the scientific litterature to adjust EVCC4 in
terms of what is learned about species sensitivity to climate change. Despite its inherrent
uncertainty it still provides interesthing results about the distribution of species’ ranges while
being a first attempt to link studies of climate change at the species level with an understanding
of climate change vulnerability at the ecosystem level. The fact that EVCC4’s theory is rooted in
many outstanding issues about our understanding of ecosystems, including the diversity-stability
relationship, that it is easy to quantify and that it can be applied practically makes it a metric
worthwhile to investigate both for ecology and conservation.
Table 2: Average EVCC4 rank per ecosystem type
Average ecosystem patch area (m2)
Average species range EVCC4 Rank Average species range
and density
1 274771.59 2850058.63
2 11219409.43 55318330.35
3 21173321.81 104512348.92
4 39363662.95 174075080.23
5 244691013.67 16272026.14
Ecosystem Vulnerability to Climate Change 42
42
0 1 2 3 4 5
11. Isla menores de 140 ha
6b. BSOT aluvial ocasionalmente inundado - Med. I
15. Sist. prod. acuático-camaronera y salina
22. Albina con escasa vegetación
16. Sabanas arbol. de graminoides cortos inund.
9. Manglares
18b. Pantanos herbáceos salobres
21. Veg. costera de trans. sobre suelos marinos muy rec.
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %)
1b. BSOT latifoliado de tierras bajas - BI
8b. B. semideciduo tropical de tierras bajas - bastante intervenido
8c. B. semideciduo tropical de tierras bajas - poco intervenido
17. Vegetación de páramo
8a. B. semideciduo tropical de tierras bajas
12. Poblados
14. Plantaciones forestales
1a. BSOT latifoliado de tierras bajas
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %)
7c. BSOT pantanoso - domin. por Campnosperma
6a. BSOT aluvial ocasionalmente inundado
2a. BSOT latifoliado submontano
7b. BSOT pantanoso - domin. por palmas
1c. BSOT latifoliado de tierras bajas - PI
2b. BSOT latifoliado submontano - BI
10. B. deciduo por la sequia, latifoliado de tierras bajas
3a. BSOT latifoliado montano
3c. BSOT latifoliado montano - PI
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria
23. Carrizales pant. y form. sim. domin. por Typha
7a. BSOT pantanoso - domin. por dicotiledoneas
5. BSOT latifoliado nuboso
18a. Pant. de cyperáceas con abund. acum. de mat. org.
2c. BSOT latifoliado submontano - PI
4a. BSOT latifoliado altimontano
20. Flujo de lava con escasa vegetación
3b. BSOT latifoliado montano - BI
4b. BSOT latifoliado altimontano - Med. I
EVCC4 rank per ecosystem type
Figure 11: Distibution of EVCC4 values for each ecosystem type
Ecosystem Vulnerability to Climate Change 43
43
0 1 2 3 4 5 6 7 8 9 10
11. Isla menores de 140 ha
6b. BSOT aluvial ocasionalmente inundado - Med. I
15. Sist. prod. acuático-camaronera y salina
22. Albina con escasa vegetación
16. Sabanas arbol. de graminoides cortos inund.
9. Manglares
18b. Pantanos herbáceos salobres
21. Veg. costera de trans. sobre suelos marinos muy rec.
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %)
1b. BSOT latifoliado de tierras bajas - BI
8b. B. semideciduo tropical de tierras bajas - bastante intervenido
8c. B. semideciduo tropical de tierras bajas - poco intervenido
17. Vegetación de páramo
8a. B. semideciduo tropical de tierras bajas
12. Poblados
14. Plantaciones forestales
1a. BSOT latifoliado de tierras bajas
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %)
7c. BSOT pantanoso - domin. por Campnosperma
6a. BSOT aluvial ocasionalmente inundado
2a. BSOT latifoliado submontano
7b. BSOT pantanoso - domin. por palmas
1c. BSOT latifoliado de tierras bajas - PI
2b. BSOT latifoliado submontano - BI
10. B. deciduo por la sequia, latifoliado de tierras bajas
3a. BSOT latifoliado montano
3c. BSOT latifoliado montano - PI
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria
23. Carrizales pant. y form. sim. domin. por Typha
7a. BSOT pantanoso - domin. por dicotiledoneas
5. BSOT latifoliado nuboso
18a. Pant. de cyperáceas con abund. acum. de mat. org.
2c. BSOT latifoliado submontano - PI
4a. BSOT latifoliado altimontano
20. Flujo de lava con escasa vegetación
3b. BSOT latifoliado montano - BI
4b. BSOT latifoliado altimontano - Med. I
Average Amphibian Rank Average Mammals Rank Average Birds Rank
Figure 12: Distribution of bird, mammal and amphibian sensitivity ranks for each ecosystem type, ranked by
total EVCC4
Ecosystem Vulnerability to Climate Change 44
44
B. Overall EVCC index
Even though justification is provided in deciding how to weigh each of the EVCCx
values, the EVCC equation was still formulated somewhat arbitrarily. As Table 3 shows, each
individual EVCC was ranked on a different scale. Species sensitivity was on a scale from 1-5,
geometry from 1-10, sea level rise from 0-14, and climatic ‘space’ was ranked from 0-19.
This scoring scheme highlights the ecosystem types with both the highest single EVCC
values and average EVCC values. Red is the highest, orange the second, and yellow the third. In
the case of a tie, all values have been assigned the same color (e.g., Max_EVCC2 has one red
rank and a ten-way tie for orange, which means there will be no yellow). A column with “-no-F”
means that it does not consider any EVCC_F values even though they have been given colors.
The table is ranked by the overall score, which was determined by adding the number of each
color-highlight for ecosystem type, where red=3, orange=2, and yellow=1 (e.g., Mangroves: 3*(2
red) + 2*(0 orange) + 1*(2 yellow) = 8), but it does not include EVCC_F because there would be
redundancies in the score. Mangroves lead this ranking scheme because they had the highest
maximum EVCC twice. The highest average overall EVCC is 30 for tropical evergreen swampy
rainforests dominated by palms, but the variance of that value is extremely high. There are only
seven patches of this ecosystem type, so the possibility of variation evening out is much lower
than many other ecosystem types. But this argument does not work for islands, the most
numerous-patched ecosystem type, which also has a high overall average EVCC variance. This
can be explained by the large range that the islands carry—the minimum and maximum EVCCs
can be very high or very low, depending on the individual ecosystem patch.
An important general observation is that the ecosystems with the highest scores are
mangroves, new coastal vegetation, swamps or marshlands, lowlands, and small islands.
Ecosystem Vulnerability to Climate Change 45
45
Table 3 Summary of individual (EVCC_1,2,3,4) and overall (EVCC_F) with statistics and rankings
Ecosystem Type
# patches
Min_EVCC1
Max_EVCC1
Ave_EVCC1
Var_EVCC1
Min_EVCC2
Max_EVCC2
Ave_EVCC2
Var_EVCC2
Min_EVCC3
Max_EVCC3
Ave_EVCC3
Var_EVCC3
Min_EVCC4
Max_EVCC4
Ave_EVCC4
Var_EVCC4
Min_EVCC_F
Max_EVCC_F
Ave_EVCC_F
Var_EVCC_F
Highlig
hts
Highlig
ht-n
o-F
RED-no-F
ORANGE-no-F
YELLOW-no-F
Score-no-F
I.A.5. Bosque de manglar 137 0 14 6.9 11.9 5 10 7.4 1.4 0 16 10.3 13.7 1 5 2.5 1.0 8 40 27.1 33.8 5 4 2 0 2 8
VI.B.3. Vegetación costera de transición sobre suelos marinos muy recientes 5 0 10 7.4 17.3 7 9 8.4 0.8 0 17 10.2 39.7 2 3 2.6 0.3 22 32 28.6 15.3 4 3 1 2 0 7
I.A.1.f.(2) Bosque siempreverde ombrofilo tropical aluvial ocasionalmente inundado 36 0 12 3.4 17.9 4 9 6.5 1.4 0 19 13.0 22.3 1 5 3.5 0.8 12 40 26.4 60.1 4 3 1 1 1 6
I.A.1.g.(2) Bosque siempreverde ombrofilo tropical pantanoso dominado por palmas 7 0 10 7.6 11.6 5 8 7.0 1.3 0 18 11.7 38.6 2 5 3.7 1.6 18 38 30.0 61.0 3 2 1 1 0 5
I.A.1.g.(3) Bosque siempreverde ombrofilo tropical pantanoso dominado por Campnosperma 8 0 8 4.6 10.6 5 8 6.1 1.0 0 18 14.0 40.9 2 5 3.3 0.8 9 35 28.0 69.7 3 2 1 1 0 5
I.A.3.a. Bosque semideciduo tropical de tierras bajas - bastante intervenido 76 0 14 2.2 16.8 4 9 6.2 1.6 0 15 8.3 8.2 1 5 2.9 1.4 10 30 19.6 25.5 2 2 1 1 0 5
I.A.3.a. Bosque semideciduo tropical de tierras bajas – poco intervenido 20 0 14 3.2 22.8 4 9 6.4 1.7 0 11 6.9 8.7 1 5 2.9 2.1 12 29 19.4 13.4 2 2 1 1 0 5
I.A.3.a. Bosque semideciduo tropical de tierras bajas 59 0 14 2.7 16.4 3 9 6.3 1.5 0 15 7.3 14.3 1 5 3.1 1.5 10 30 19.3 16.1 2 2 1 1 0 5
Isla menores de 140 hectareas 530 0 14 6.0 33.4 6 9 6.3 0.3 0 18 0.8 8.7 1 5 1.6 1.3 7 38 14.7 41.6 3 3 0 2 0 4
V.D.1.a. Pantanos de cyperáceas con abundante acumulación de material orgánico 3 0 0 0.0 0.0 6 7 6.7 0.3 11 18 13.3 16.3 3 5 4.3 1.3 23 27 24.3 5.3 2 2 0 2 0 4
SP.A. Sistema productivo con vegetación leñosa natural o espontánea significativa (10-50 %) 60 0 10 3.2 15.7 4 9 6.3 1.3 0 18 8.1 15.9 1 5 3.2 1.7 8 37 20.9 43.1 2 2 0 2 0 4
P. Poblados 21 0 14 3.9 27.8 5 8 6.8 0.7 0 15 8.6 20.8 1 5 3.1 2.0 16 40 22.4 39.5 2 1 1 0 0 3
I.A.1.d.(1) Bosque siempreverde ombrofilo tropical latifoliado altimontano (1500-2000 m Car, 1800-2300m Pac) - medianamente interv. 1 0 0 0.0 0.0 5 5 5.0 0.0 3 3 3.0 0.0 5 5 5.0 0.0 13 13 13.0 0.0 1 1 1 0 0 3
SP.B. Sistema productivo con vegetación leñosa natural o espontánea significativa (<10 %) 46 0 12 4.6 16.0 3 9 6.4 1.7 0 15 9.2 16.3 1 5 2.7 1.6 9 34 22.8 29.6 1 1 0 1 0 2
I.A.1.a.(1) Bosque siempreverde ombrofilo tropical latifoliado de tierras bajas - bastante intervenido 45 0 12 3.5 18.1 4 9 6.5 1.3 3 15 8.9 9.8 1 5 2.8 1.8 12 36 21.8 41.2 1 1 0 1 0 2
I.A.1.a.(1) Bosque siempreverde ombrofilo tropical latifoliado de tierras bajas 77 0 12 3.4 16.0 2 9 6.1 1.7 0 17 8.7 15.0 1 5 3.2 2.1 11 32 21.5 26.4 1 1 0 1 0 2
VI.D. Albina con escasa vegetación 1 3 3 3.0 0.0 8 8 8.0 0.0 8 8 8.0 0.0 2 2 2.0 0.0 21 21 21.0 0.0 1 1 0 1 0 2
I.A.1.c.(1) Bosque siempreverde ombrofilo tropical latifoliado montano (1000-1500m Caribe, 1200-1800 m Pacífico) - bastante intervenido 14 0 0 0.0 0.0 5 7 6.0 0.8 3 13 6.6 8.7 3 5 4.6 0.4 12 25 17.1 11.1 1 1 0 1 0 2
Ecosystem Vulnerability to Climate Change 46
46
VI.A.d. Flujo de lava con escasa vegetación 4 0 0 0.0 0.0 5 7 5.8 0.9 6 14 9.8 18.9 4 5 4.5 0.3 16 23 20.0 12.7 1 1 0 0 1 1
I.A.1.d.(1) Bosque siempreverde ombrofilo tropical latifoliado altimontano (1500-2000 m Caribe, 1800-2300 m Pacífico) 4 0 0 0.0 0.0 5 8 6.0 2.0 3 12 6.5 19.0 4 5 4.5 0.3 13 22 17.0 15.3 1 1 0 0 1 1
I.A.1.b.(1) Bosque siempreverde ombrofilo tropical latifoliado submontano (500-1000 m Caribe, 700-1200 m Pacífico) - poco intervenido 2 0 0 0.0 0.0 4 6 5.0 2.0 5 6 5.5 0.5 4 5 4.5 0.5 13 17 15.0 8.0 1 1 0 0 1 1
I.A.1.g.(1) Bosque siempreverde ombrofilo tropical pantanoso dominado por dicotiledóneas 5 0 9 3.0 18.0 6 8 7.0 0.5 9 11 9.6 0.8 3 5 4.2 0.7 20 32 23.8 25.2 0 0 0 0 0 0
I.A.1.f.(2)(a) Bosque siempreverde ombrofilo tropical aluvial, ocasionalmente inundado dominado por Prioria 6 0 0 0.0 0.0 5 8 6.2 1.4 9 14 12.2 4.6 3 5 4.0 0.4 18 26 22.3 7.5 0 0 0 0 0 0
I.A.1.a.(1) Bosque siempreverde ombrofilo tropical latifoliado de tierras bajas - poco intervenido 32 0 12 1.7 11.1 4 8 6.3 1.4 5 16 9.4 11.0 1 5 3.8 2.1 15 32 21.2 26.1 0 0 0 0 0 0
V.C.2.b. Vegetación de páramo 2 0 0 0.0 0.0 5 6 5.5 0.5 12 13 12.5 0.5 3 3 3.0 0.0 21 21 21.0 0.0 0 0 0 0 0 0
SP.C. Sistema productivo acuático-camaronera y salina 6 0 12 4.7 21.9 5 8 6.8 1.4 0 11 7.3 14.3 2 2 2.0 0.0 19 25 20.8 5.8 0 0 0 0 0 0
V.E.1.a.2. Pantanos herbáceos salobres 4 0 8 2.0 16.0 6 8 7.0 0.7 7 12 9.3 4.3 2 4 2.5 1.0 17 31 20.8 46.9 0 0 0 0 0 0
I.B.1.a.(1) Bosque deciduo por la sequía, latifoliado de tierras bajas 5 0 8 2.8 15.2 4 7 5.8 1.7 6 10 8.2 2.2 3 5 3.8 0.7 17 29 20.6 24.3 0 0 0 0 0 0
SP.B.1. Plantaciones forestales 6 0 0 0.0 0.0 6 7 6.3 0.3 7 15 10.8 9.0 2 4 3.2 1.0 17 23 20.3 4.7 0 0 0 0 0 0
V.A.2.d. Sabanas arboladas de graminoides cortos inundables 5 0 6 1.2 7.2 6 8 7.0 0.5 7 12 9.6 3.3 2 4 2.4 0.8 19 23 20.2 3.2 0 0 0 0 0 0
VII.B. Carrizales pantanosos y formaciones similares principalmente de Typha 5 0 0 0.0 0.0 4 7 5.4 1.8 9 11 10.0 1.0 4 4 4.0 0.0 17 21 19.4 3.3 0 0 0 0 0 0
I.A.1.f.(2) Bosque siempreverde ombrofilo tropical aluvial ocasionalmente inundado - medianamente intervenido 1 0 0 0.0 0.0 5 5 5.0 0.0 11 11 11.0 0.0 2 2 2.0 0.0 18 18 18.0 0.0 0 0 0 0 0 0
I.A.1.c.(1) Bosque siempreverde ombrofilo tropical latifoliado montano (1000-1500 m Caribe, 1200-1800 m Pacífico) 18 0 0 0.0 0.0 4 8 5.6 1.3 3 14 8.2 14.9 2 5 3.9 1.3 10 24 17.8 16.8 0 0 0 0 0 0
I.A.1.e.(1) Bosque siempreverde ombrofilo tropical latifolado nuboso (2000-3000 m Caribe, 2300-3000 m Pacífico) 3 0 0 0.0 0.0 6 6 6.0 0.0 5 9 7.3 4.3 3 5 4.3 1.3 14 20 17.7 10.3 0 0 0 0 0 0
I.A.1.b.(1) Bosque siempreverde ombrofilo tropical latifoliado submontano (500-1000 m Caribe, 700-1200 m Pacífico) - bastante interv. 21 0 0 0.0 0.0 3 8 5.6 1.5 4 13 8.0 10.6 2 5 3.8 1.8 12 22 17.4 11.4 0 0 0 0 0 0
I.A.1.b.(1) Bosque siempreverde ombrofilo tropical latifoliado submontano (500-1000 m Caribe, 700-1200 m Pacífico) 27 0 0 0.0 0.0 3 8 5.7 1.4 2 14 7.6 11.2 2 5 3.5 1.1 9 24 16.9 13.5 0 0 0 0 0 0
I.A.1.c.(1) Bosque siempreverde ombrofilo tropical latifoliado montano (1000-1500 m Caribe, 1200-1800 m Pacífico) - poco intervenido 1 0 0 0.0 0.0 5 5 5.0 0.0 6 6 6.0 0.0 4 4 4.0 0.0 15 15 15.0 0.0 0 0 0 0 0 0
Ecosystem Vulnerability to Climate Change 47
47
Figure 13 Distribution of average overall EVCC values for each ecosystem type
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
11. Isla menores de 140 ha
2c. BSOT latifoliado submontano - PI
3c. BSOT latifoliado montano - PI
2a. BSOT latifoliado submontano
4a. BSOT latifoliado altimontano
3b. BSOT latifoliado montano - BI
2b. BSOT latifoliado submontano - BI
5. BSOT latifoliado nuboso
3a. BSOT latifoliado montano
6b. BSOT aluvial ocasionalmente inundado - Med. I
8a. B. semideciduo tropical de tierras bajas
8c. B. semideciduo tropical de tierras bajas - poco intervenido
23. Carrizales pant. y form. sim. domin. por Typha
8b. B. semideciduo tropical de tierras bajas - bastante intervenido
20. Flujo de lava con escasa vegetación
16. Sabanas arbol. de graminoides cortos inund.
14. Plantaciones forestales
10. B. deciduo por la sequia, latifoliado de tierras bajas
18b. Pantanos herbáceos salobres
15. Sist. prod. acuático-camaronera y salina
13a. Sist. prod. con veg. leñosa nat. o esp. signif. (10-50 %)
17. Vegetación de páramo
22. Albina con escasa vegetación
1c. BSOT latifoliado de tierras bajas - PI
1a. BSOT latifoliado de tierras bajas
1b. BSOT latifoliado de tierras bajas - BI
6c. BSOT aluvial ocasionalmente inundado - domin. por Prioria
12. Poblados
13b. Sist. prod. con veg. leñosa nat. o esp. signif. (<10 %)
7a. BSOT pantanoso - domin. por dicotiledoneas
18a. Pant. de cyperáceas con abund. acum. de mat. org.
6a. BSOT aluvial ocasionalmente inundado
9. Manglares
7c. BSOT pantanoso - domin. por Campnosperma
21. Veg. costera de trans. sobre suelos marinos muy rec.
7b. BSOT pantanoso - domin. por palmas
EVCC rank per ecosystem type
Ecosystem Vulnerability to Climate Change 48
48
Map 5
Ecosystem Vulnerability to Climate Change 49
49
C. Applications of the index
Only the overall EVCC values are compared to the three applications. Relationships
between the EVCC index and the applications are provided and discussed in each section.
1) Degree of human intervention
Before comparing the EVCC to this measurement, the percent area of each intervention
level was calculated (Table 4).
Table 4 Percentage of human intervention
Level of Intervention Total area of intervention type (ha)
% area of Panama
0 None documented 3096464.7364 41.58%
1 Natural system with low intervention 187260.6452 2.51%
2 Natural system with medium intervention 9389.7263 0.13%
3 Natural system with high intervention in mountains 130335.2176 1.75%
4 Natural system with high intervention in lowlands 1107875.1517 14.88%
5 Productive system with 10-50% natural veg 1739669.8057 23.36%
6 Productive system with <10% natural veg 1142648.6367 15.34%
7 Agroforestry 5267.1732 0.07%
8 Shrimp / Salt production 7391.1249 0.10%
9 Populated place 21170.0424 0.28%
All area of Panama 7447472.2601 100.00%
These results are based solely on UNESCO’s descriptions of ecosystem types and their
delineations on its ecosystem map. The three mid-levels (4, 5, and 6) of intervention account for
over half of the land in Panama. The areas with the highest levels of intervention (agroforestry,
shrimp and salt production, and populated places) cover less than one-half percent of the land,
but are should not be deemed insignificant because of the strong human interest placed in these
systems (e.g., ecosystem services).
Ecosystem Vulnerability to Climate Change 50
50
Table 5 EVCC values compared to Degrees of Intervention
Level of intervention Avg_EVCC1 Ave_EVCC2 Ave_EVCC3 Ave_EVCC4 Ave_EVCC
0 none 5.0987 6.4086 4.6817 2.2658 18.4548
1
Natural system with low intervention 2.1636 6.2727 8.2545 3.4727 20.1636
2
Natural system with medium intervention 0.0000 5.0000 7.0000 3.5000 15.5000
3
Natural system with high intervention in mountains 0.0000 5.7714 7.4286 4.0857 17.2857
4
Natural system with high intervention in lowlands 2.7025 6.3223 8.5207 2.8678 20.4132
5
Productive system with 10-50% natural veg 3.2333 6.3167 8.1333 3.2333 20.9167
6
Productive system with <10% natural veg 4.5652 6.3696 9.1522 2.7391 22.8261
7 Agroforestry 0.0000 6.3333 10.8333 3.1667 20.3333
8 Shrimp / Salt production 4.6667 6.8333 7.3333 2.0000 20.8333
9 Populated place 3.9048 6.7619 8.5714 3.1429 22.3810
Red = highest average; orange = second highest; yellow = third highest.
Blue = most commonly ranked with high vulnerability
Table 5 demonstrates the individual and overall average EVCC values for each level of
intervention. Notable levels are zero, six, eight, and nine (highlighted in beige) because of the
frequency of high vulnerabilities (red, orange, and yellow). The zero intervention level includes
all islands less than 140 hectares, of which there are over 530 ecosystem patches (Appendix 1).
That average EVCC1 is the highest here makes sense because so many islands are susceptible to
flooding due to sea level rise. They also receive a high EVCC2a values because often islands
have no “edge” in the edge to core ratio. Intervention level six, productive systems with less
than 10% natural or spontaneous vegetation-type ecosystems received the highest overall
average EVCC. Except for EVCC4 (which makes sense because this is a highly cultivated
ecosystem type that has likely removed native species and decreased biodiversity), all EVCC1,2,3
Ecosystem Vulnerability to Climate Change 51
51
are relatively high. The fact that this study has ranked this intervention level as the second most
vulnerable according to climatic ‘space’ is worth attention. Farming relies upon climatic cycles
that have been learned and adapted to for centuries, and the geographical locations of the patches
of this agricultural ecosystem type are such that climate change projections illustrate some of the
highest changes in temperature and precipitation. Intervention level eight, shrimp and/or salt
production, can mostly be explained by the combination of their close proximity to coastlines
and the almost absent slope of the land. The farming of shrimp and salt requires vast extents of
flat land that are occasionally inundated by sea water at the highest high tides of the year. It is
obvious that sea level rise will greatly impact these areas. Perhaps the most worrisome to
policymakers is that the populated places have received the second highest overall average
vulnerability. It is in the upper half of sea level rise vulnerability and in the lower half of species
sensitivity. The pertinence of its high EVCC2 ranking is questionable because it is not one of the
natural systems, which are more susceptible to ramifications of stochastic processes when they
have irregular shapes or many edges to welcome the perturbations. This human-dominated
system is very buffered by these types of effects. It is also notable that the projected changes in
temperature and precipitation have more of an impact here than other places.
2) Protected areas
The percentage of area protected in each ecosystem type is compared to its mean EVCC value in
(Figure 83). This shows that some of the most vulnerable ecosystem types are also the ones with
the least protection on average. Also, the average EVCC per area (at a resolution of 1 km2) for
the overall EVCC and the four EVCC components was calculated for all areas that are protected,
all areas that are not protected and the whole of Panama (see Table 6). This allows to evaluate
Ecosystem Vulnerability to Climate Change 52
52
how the current protected area network accounts for ecosystem vulnerability to climate change
according to the EVCC. The mean EVCC should be higher within protected areas if the network
generally protects more vulnerable regions. If lower EVCC means that areas outside of the
network are generally more vulnerable to climate change (according to the EVCC) and thus, that
the protected areas network is inadequate in protecting vulnerable areas. We found the latter to
be the case for the overall EVCC, EVCC1 and EVCC3, but the effect is much more pronounced
in EVCC1. Nonetheless, the fact that in almost no case is the average EVCC in protected areas
higher means that the current network of reserve is inadequate to protect the vulnerable areas of
Panama in terms of future climate change. Additionally, something to keep in mind with this
analysis is that protected marine areas are excluded because the ecosystem patch map does not
include aquatic ecosystems.
Table 6: Average EVCCs per area in Panama as a whole
and inside and outside of protected areas
inside
PAs
outside
PAs Panama
EVCC 17.63062 17.787722 17.842452
EVCC1 1.34952 1.978375 2.197648
EVCC2 6.384813 6.094201 5.992869
EVCC3 5.129844 5.132692 5.133686
EVCC4 4.766585 4.582453 4.518249
Comparative Analysis: Protected Areas and EVCC for Ecosystem Type
0
5
10
15
20
25
30
35
7b. B
SOT pantanoso - d
omin. por p
almas
21. V
eg. costera de tra
ns. sobre suelos marinos muy re
c.
7c. B
SOT pantanoso - d
omin. por C
ampnosperm
a
9. M
anglares
6a. B
SOT aluvial ocasionalmente inundado
18a. P
ant. d
e cyperáceas con abund. acum. de mat. o
rg.
7a. B
SOT pantanoso - d
omin. por d
icotile
doneas
13b. S
ist. p
rod. con veg. le
ñosa nat. o
esp. signif. (<
10 %
)
12. P
oblados
6c. B
SOT aluvial ocasionalmente inundado - d
omin. por P
rioria
1b. B
SOT latifo
liado de tie
rras bajas - B
I
1a. B
SOT latifo
liado de tie
rras bajas
1c. B
SOT latifo
liado de tie
rras bajas - P
I
17. V
egetación de páramo
22. A
lbina con escasa vegetación
13a. S
ist. p
rod. con veg. le
ñosa nat. o
esp. signif. (1
0-50 %
)
15. S
ist. p
rod. acuático-camaronera y salina
18b. P
antanos herbáceos salobres
10. B
. deciduo por la
sequia, la
tifoliado de tie
rras bajas
14. P
lantaciones forestales
16. S
abanas arbol. d
e graminoides corto
s inund.
20. F
lujo de lava con escasa vegetación
8b. B
. semideciduo tro
pical de tie
rras bajas - b
astante intervenido
23. C
arriz
ales pant. y
form
. sim. domin. por T
ypha
8c. B
. semideciduo tro
pical de tie
rras bajas - p
oco intervenido
8a. B
. semideciduo tro
pical de tie
rras bajas
6b. B
SOT aluvial ocasionalmente inundado - M
ed. I
3a. B
SOT latifo
liado montano
5. B
SOT latifo
liado nuboso
2b. B
SOT latifo
liado submontano - B
I
3b. B
SOT latifo
liado montano - B
I
4a. B
SOT latifo
liado altim
ontano
2a. B
SOT latifo
liado submontano
2c. B
SOT latifo
liado submontano - P
I
3c. B
SOT latifo
liado montano - P
I
11. Is
la menores de 140 ha
4b. B
SOT latifo
liado altim
ontano - M
ed. I
Ecosystem Type
Average EVCC
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percentage Protected
Ave_EVCC
Percent Protected
Figure 8 Protected Areas Analysis
Ecosystem Vulnerability to Climate Change 54
54
3) Species richness and endemism
Figure 94: Comparing the overall EVCC with patterns of species richness and endemism in Panama
Species richness and endemicity were calculated for each cell of a 10 x 10 km grid of
Panama (right part of Figure 14). It should be noted that endemism for mammals, birds, and
amphibians is inversely related to the species richness in Panama. That is, there are more birds
than mammals and more mammals than amphibians in the country, but in terms of endemism,
there are the most amphibians and the least birds. For each cell, the overall EVCC was compared
with a map of species richness and endemism to see if any patterns could be extracted or whether
areas of high biodiversity also happened to be more vulnerable to climate change according to
the EVCC (Figure 14). Generally no correlation were found for neither species richness nor
endemism excepting for EVCC3 where regions of high endemism had a low score (this is
vvss..
Ecosystem Vulnerability to Climate Change 55
55
reflected in the overall EVCC as well) (Figure 15). In order to highlight regions that had both
high species richness and overall EVCC, species richness per ecosystem patch was multiplied by
overall EVCC score to yield Map 6. The region that stands out the most in this map is Bocas del
Toro, but the west of Panama, the areas of the Canal and the south of the Darien are highlighted
as well.
Figure 10: Correlation of species richness (a-e) and endemism (f-j) with EVCCs for each cell of a 10x10 km
grid of Panama
Ecosystem Vulnerability to Climate Change 56
56
Map 6 Combining overall EVCC with species richness
Ecosystem Vulnerability to Climate Change 57
57
VI. Conclusions
The objective of this project was to conduct a prelimary assesment of ecosystem vulnerability to
climate change in Panama. The final results include both a set of maps from which spatial
information (including specific information on ecosystem patches and trends) can be extracted
and an analysis of the EVCC in turns of applications relevant to conservation. These were
communicated in a comprehensive format to the host institution as both a hard copy (including a
copy of important GIS data and maps) and a presentation. This project can not only be valuable
for the conservation of biodiversity of Panama both also to other regions by providing a
framework that can be followed to build EVCC assessements in other geographical locations.
The EVCC is a preliminary assessment and is far from being perfect. There are a number
of issues that could be adressed to improve it and a number of uncertainties that need to be
accounted for when applying the EVCC in policy-making. We outline some of them here.
First, certain types of data were unavailable at the time of the assessment but should
preferably be included in future assessments if possible. These include other scenarios of the
CCM than the “business-as-usual” one, higher climate resolution for the CCM and tree
distributions. Moreover, islands are underrated in this assessment because of overlaying issues in
EVCC1 (such that an island’s overall elevation is misrepresented) and because many of them
were not included in EVCC3 since they are so small. Also, future climate for islands is harder to
predict in the future because of micro-climate issues brought by the surrounding water masses.
Finally, the effects of the ranking scheme on the overall EVCC should be investigated. While
efforts were put into using the most relevant scheme for each data type, it might well be that
these had unintended effects on the overall distribution of the EVCC.
Ecosystem Vulnerability to Climate Change 58
58
Second, the EVCC can only be as good as the data it is based on is. There are a number
of uncertainties in the accuracy of the data and thus, the results are not definitive. Possible
problems of accuracies include the digital elevation model of sea level rise vulnerability,
predictions of future climate and species distributions. There are also uncertainties associated
with the theories behind the EVCC. Considerable efforts were put into including the latest
scientific consensys on the effects of climate change on species and ecosystems but these are
always evolving. Likewise, it might well be that the four components selected to build the EVCC
are not enough, or too much, and this should be investigated in light of the most recent scientific
litterature. In general thus, studies like this done in GIS should be used as indicators of where
further research or conservation efforts should be directed. These results cannot replace on-sight
situation analyses, but at least such projects can suggest locations of high concern. It is essential
to use a comprehensive framework to adress uncertainty in data and theory, especially in areas
like climate change where these uncertainties are high.
To conclude, EVCC has the potential to be a very interesthing tool for conservation and
ecology. It is very much a work-in-progress and it has been structured in a flexible way so that it
can be improved easily. One of the biggest challenges to use the EVCC in conservation will be to
decide what its relative importance is compared to other human drivers of change in biodiversity
like habitat loss, over-exploitation and invasive species. Nonewithstanding all the uncertainties
associated with the EVCC, it is the first attempt at qualifying vulnerability to climate change at
the ecosystem level and sets up a template that can be used directly or adjusted according ton
specific needs. Because it estimates an important threat to ecosystems, climate change, which is
usually over-looked in biodiversity assesments, it can become, if properly applied, a very useful
tool for conservation.
Ecosystem Vulnerability to Climate Change 59
59
X. References
Broenniman O, Thuiller W, Hughes G, et al. (2006) Do geographic distribution, niche property
and life form explain plants' vulnerability to global change? Global Change Biology 12,
1079-1093.
Ducklow HW, Baker K, Martinson DG, et al. (2007) Marine pelagic ecosystems: The West
Antarctic Peninsula. Philosophical Transactions of the Royal Society B-Biological
Sciences 362, 67-94.
Frederiksen M, Edwards M, Richardson AJ, Halliday NC, Wanless S (2006) From plankton to
top predators: bottom-up control of a marine food web across four trophic levels. Journal
of Animal Ecology 75, 1259-1268.
Hampe A (2004) Bioclimate envelope models: what they detect and what they hide. Global
Ecology and Biogeography 13, 469-471.
Hawkins BA, Field R, Cornell HV, et al. (2003) Energy, water, and broad-scale geographic
patterns of species richness. Ecology 84, 3105-3117.
Hooper DU, Chapin FS, Ewel JJ, et al. (2005) Effects of biodiversity on ecosystem functioning:
A consensus of current knowledge. Ecological Monographs 75, 3-35.
Ibanez I, Clark JS, Dietze MC, et al. (2006) Predicting biodiversity change: Outside the climate
envelope, beyond the species-area curve. Ecology 87, 1896-1906.
Institute WR (2007) EarthTrends: Environmental Information. Available at
http://earthtrends.wri.org. World Resources Institute, Washington DC.
IPCC (2001) Climate change 2001: impacts, adaptations and vulnerability. Contribution of
Working Group II to the Third Assesment Report of the Intergovernmental Panel on
Climate Change (IPCC). Cambridge University Press, Cambridge.
IPCC (2007) Climate Change 2007: The Physical Science Basis - Summary for Policymakers,
Geneva.
IUCN (2006) 2006 IUCN Red List of Threatened Species. IUCN-SSC
(http://www.iucnredlist.org).
Janzen DH (1983) No Park Is an Island - Increase in Interference from Outside as Park Size
Decreases. Oikos 41, 402-410.
Jiguet F, Julliard R, Thomas CD, et al. (2006) Thermal range predicts bird population resilience
to extreme high temperatures. Ecology Letters 9, 1321-1330.
McGill University (2007) Academic Integrity – Researchers. Available at
http://www.mcgill.ca/integrity/researchers/
Metzger MJ, Leemans R, Schroter D (2005) A multidisciplinary multi-scale framework for
assessing vulnerabilities to global change. International Journal of Applied Earth
Observation and Geoinformation 7, 253-267.
Murcia C (1995) Edge Effects in Fragmented Forests - Implications for Conservation. Trends in
Ecology & Evolution 10, 58-62.
Nature Serve, 2007. InfoNatura website (for species lists and endemism). Available at
www.natureserve.org
Parmesan C (2006) Ecological and evolutionary responses to recent climate change. Annual
Review of Ecology Evolution and Systematics 37, 637-669.
Pearson RG (2006) Climate change and the migration capacity of species. Trends in Ecology &
Evolution 21, 111-113.
Ecosystem Vulnerability to Climate Change 60
60
Peterson G, Allen CR, Holling CS (1998) Ecological resilience, biodiversity, and scale.
Ecosystems 1, 6-18.
Pounds JA, Bustamante MR, Coloma LA, et al. (2006) Widespread amphibian extinctions from
epidemic disease driven by global warming. Nature 439, 161-167.
Sala OE, Chapin FS, Armesto JJ, et al. (2000) Biodiversity - Global biodiversity scenarios for
the year 2100. Science 287, 1770-1774.
Saunders DA, Hobbs RJ, Margules CR (1991) Biological Consequences of Ecosystem
Fragmentation - a Review. Conservation Biology 5, 18-32.
Scholze M, Knorr W, Arnell NW, Prentice IC (2006) A climate-change risk analysis for world
ecosystems. Proceedings of the National Academy of Sciences of the United States of
America 103, 13116-13120.
Schroter D, Cramer W, Leemans R, et al. (2005) Ecosystem service supply and vulnerability to
global change in Europe. Science 310, 1333-1337.
Shafer CL (1999) National park and reserve planning to protect biological diversity: some basic
elements. Landscape and Urban Planning 44, 123-153.
Sutherland WJ (2006) Predicting the ecological consequences of environmental change: a review
of the methods. Journal of Applied Ecology 43, 599-616.
Thomas CD, Cameron A, Green RE, et al. (2004) Extinction risk from climate change. Nature
427, 145-148.
Thuiller W, Lavorel S, Araujo MB (2005a) Niche properties and geographical extent as
predictors of species sensitivity to climate change. Global Ecology and Biogeography 14,
347-357.
Thuiller W, Lavorel S, Araujo MB, Sykes MT, Prentice IC (2005b) Climate change threats to
plant diversity in Europe. Proceedings of the National Academy of Sciences of the United
States of America 102, 8245-8250.
Thuiller W, Midgley GF, Hughes GO, et al. (2006) Endemic species and ecosystem sensitivity to
climate change in Namibia. Global Change Biology 12, 759-776.
Walther GR, Post E, Convey P, et al. (2002) Ecological responses to recent climate change.
Nature 416, 389-395.
Ecosystem Vulnerability to Climate Change 61
61
Appendix 1. Ecosystem Types and Frequencies
UNESCO ecosystem type
# of
patches Total area (ha)
I.A.1.a.(1) Bosque siempreverde ombrofilo tropical latifoliado de tierras bajas 77 1068489.0140
I.A.1.a.(1) Bosque siempreverde ombrofilo tropical latifoliado de tierras bajas -
bastante intervenido 45 654997.4415
I.A.1.a.(1) Bosque siempreverde ombrofilo tropical latifoliado de tierras bajas -
poco intervenido 32 118382.9463
I.A.1.b.(1) Bosque siempreverde ombrofilo tropical latifoliado submontano (500-
1000 m Caribe, 700-1200 m Pacífico) 27 679494.4446
I.A.1.b.(1) Bosque siempreverde ombrofilo tropical latifoliado submontano (500-
1000 m Caribe, 700-1200 m Pacífico) - bastante intervenido 21 84856.4851
I.A.1.b.(1) Bosque siempreverde ombrofilo tropical latifoliado submontano (500-
1000 m Caribe, 700-1200 m Pacífico) - poco intervenido 2 4747.5028
I.A.1.c.(1) Bosque siempreverde ombrofilo tropical latifoliado montano (1000-
1500 m Caribe, 1200-1800 m Pacífico) 18 189520.8868
I.A.1.c.(1) Bosque siempreverde ombrofilo tropical latifoliado montano (1000-
1500 m Caribe, 1200-1800 m Pacífico) - bastante intervenido 14 45478.7325
I.A.1.c.(1) Bosque siempreverde ombrofilo tropical latifoliado montano (1000-
1500 m Caribe, 1200-1800 m Pacífico) - poco intervenido 1 5093.5239
I.A.1.d.(1) Bosque siempreverde ombrofilo tropical latifoliado altimontano
(1500-2000 m Caribe, 1800-2300 m Pacífico) 4 38690.5909
I.A.1.d.(1) Bosque siempreverde ombrofilo tropical latifoliado altimontano
(1500-2000 m Caribe, 1800-2300 m Pacífico) - medianamente intervenido 1 6594.3574
I.A.1.e.(1) Bosque siempreverde ombrofilo tropical latifoliado nuboso (2000-
3000 m Caribe, 2300-3000 m Pacífico) 3 21339.9920
I.A.1.f.(2) Bosque siempreverde ombrofilo tropical aluvial ocasionalmente
inundado 36 77392.5517
I.A.1.f.(2) Bosque siempreverde ombrofilo tropical aluvial ocasionalmente
inundado - medianamente intervenido 1 2795.3689
I.A.1.f.(2)(a) Bosque siempreverde ombrofilo tropical aluvial, ocasionalmente
inundado dominado por Prioria 6 13410.2662
I.A.1.g.(1) Bosque siempreverde ombrofilo tropical pantanoso dominado por
dicotiledóneas 5 4124.8543
I.A.1.g.(2) Bosque siempreverde ombrofilo tropical pantanoso dominado por
palmas 7 3760.6590
I.A.1.g.(3) Bosque siempreverde ombrofilo tropical pantanoso dominado por
Campnosperma 8 23960.6124
I.A.3.a. Bosque semideciduo tropical de tierras bajas 59 708492.0301
I.A.3.a. Bosque semideciduo tropical de tierras bajas - bastante intervenido 76 452877.7101
I.A.3.a. Bosque semideciduo tropical de tierras bajas - poco intervenido 20 59036.6721
I.A.5. Bosque de manglar 137 212650.1390
I.B.1.a.(1) Bosque deciduo por la sequía, latifoliado de tierras bajas 5 7059.7005
Isla menores de 140 hectáreas 530 8291.8601
P. Poblados 21 21170.0424
SP.A. Sistema productivo con vegetación leñosa natural o espontánea
significativa (10-50 %) 60 1739669.8057
SP.B. Sistema productivo con vegetación leñosa natural o espontánea
significativa (<10 %) 46 1142648.6367
SP.B.1. Plantaciones forestales 6 5267.1732
SP.C. Sistema productivo acuático-camaronera y salina 6 7391.1249
Ecosystem Vulnerability to Climate Change 62
62
V.A.2.d. Sabanas arboladas de graminoides cortos inundables 5 4329.8005
V.C.2.b. Vegetación de páramo 2 2484.9664
V.D.1.a. Pantanos de cyperáceas con abundante acumulación de material orgánico 3 1305.0761
V.E.1.a.2. Pantanos herbáceos salobres 4 2002.2752
VI.A.d. Flujo de lava con escasa vegetación 4 5987.8219
VI.B.3. Vegetación costera de transición sobre suelos marinos muy recientes 5 3855.5363
VI.D. Albina con escasa vegetación 1 874.2372
VII.B. Carrizales pantanosos y formaciones similares principalmente de Typha 5 18385.8150
TOTAL 1303 ~7446911
Ecosystem Vulnerability to Climate Change 63
63
Appendix 2. Endemic Species in Panama
Scientific Name Common Name Range (hectares)
Mammalia
Alouatta coibensis Coiba Island Howling Monkey 782260
Bassaricyon pauli Chiriqui Olingo 139384
Bradypus pygmaeus Escudo Island Three-toed Sloth 341
Coendou rothschildi Rothschild’s Porcupine 5829280
Cryptotis endersi Ender's Small-eared Shrew 142297
Cryptotis mera Darien Small-eared Shrew 645692
Dasyprocta coibae Coiban Agouti ?
Dermanura incomitata Isla Escudo Fruit-eating Bat 341
Isthmomys flavidus Yellow Isthmus Rat 647036
Liomys adspersus Panamanian Spiny Pocket Mouse 2636778
Marmosops invictus Slaty Slender Mouse Opossum 2064606
Neacomys pictus Painted Bristly Mouse 322680
Orthogeomys dariensis Darien Pocket Gopher 1832796
Rhipidomys scandens Mount Pirri Climbing Mouse 215628
Tylomys fulviventer A Climbing Rat 10469
Tylomys panamensis Panama Climbing Rat 359135
Aves
Anthracothorax veraguensis Veraguan Mango 3044498
Chlorospingus inornatus Pirre Bush-Tanager 98657
Leptotila battyi Azuero Dove 530265
Margarornis bellulus Beautiful Treerunner 48672
Piculus callopterus Stripe-cheeked Woodpecker 538297
Phylloscartes flavovirens Panama Tyrannulet 1727088
Pselliophorus luteoviridis Yellow-green Finch 121165
Selasphorus ardens Glow-throated Hummingbird 228079
Amphibia
Atelopus certus 10510
Atelopus limosus 44119
Atelopus zeteki Golden Frog 158879
Bolitoglossa anthracina 111536
Bolitoglossa cuna 38846
Bolitoglossa taylori 87142
Bufo peripatetes 12818
Caecilia volcani 700454
Dendrobates arboreus 22822
Dendrobates claudiae 6940
Dendrobates speciosus 37580
Dendrobates vicentei 153146
Eleutherodactylus azueroensis 150366
Eleutherodactylus emcelae 175598
Eleutherodactylus jota 5850
Ecosystem Vulnerability to Climate Change 64
64
Eleutherodactylus laticorpus 28252
Eleutherodactylus monnichorum 13670
Eleutherodactylus museosus 336100
Eleutherodactylus punctariolus 379358
Epipedobates maculates ?
Hyla graceae 187471
Hyla infucata 4965
Hyla thysanota 6550
Oedipina maritime 182
Oscaecilia elongate 5062
Pipa myersi 63184
Rana pipiens Northern Leopard Frog 445030
Rana sp. 2 Common Leopard Frog ?
Scinax altae 1251113
AREA of PANAMA ~7446911
Note: One endemic mammal and two endemic amphibians do not have a range attributed to them
because there are incongruencies between the online database and the Arc files. These
three species have not been recorded in the GIS and therefore cannot be included in the
EVCC4. This is possibly because the online database is more up to date than the Arc
files.
Ecosystem Vulnerability to Climate Change 65
65
Appendix 3. Special Focus Maps
Map 7 Sea Level
Rise Focus: Bocas
del Toro. This
display of one of
the regions in
Panama with some
of the highest
EVCC1 values
shows small
islands, tropical
evergreen swampy
rainforests
dominated by
palms, and
mangroves.
Map 8 Overall
EVCC Focus:
Canal Zone. The
Colón area
frequently has high-
vulnerability
ecosystem patches
of the types:
mangroves and
populated places.
Ecosystem Vulnerability to Climate Change 66
66
Map 9 Overall
EVCC Focus:
Golfo de Chiriquí.
The high-
vulnerability
patches here are
commonly lowland
tropical rainforests,
fairly intensive
agricultural areas,
as well as more
small islands and
mangroves.
Ecosystem Vulnerability to Climate Change 67
67
Appendix 4. Chronogram of Activities & Time Spent
Dates, times, descriptions Total person-
hours spent on
internship
4 January (17:00-19:00)
• Internship Cocktail – introductions and visit to CATHALAC
0
11 & 12 January (8:30-16:30)
• Orientation to CATHALAC and SERVIR with Milton Solano
32
13 January (20:00-20:30)
• Informal meeting with Dr. Roberto Ibáñez to explain preliminary ideas of
the project
1
15 January (19:30-20:30)
• Meeting with supervisor Emil Cherrington to start refining the study
question: “What areas are vulnerable to climate change?”
2
18 & 19 January (8:30-16:30)
• Preliminary requests for data required for the project
• Meeting with Noel Trejos (specialist in Integrated Management of Water
Resources and Community Development) to discuss possible field work
(18 Jan, 13:00)
• Video-meeting with SERVIR Project Manager from NASA Dan Irwin (18
Jan, 15:00)
• Meeting with Joel Peréz Fernandez (specialist in Weather and Climate
Technology) to learn about climatological data and models (19 Jan, 15:00)
32
25 & 26 January (8:30-16:30)
• Preliminary literature review on climate change’s effects on biodiversity
• Overview of spatial analysis, algorithm, and interpolation capabilities of
ArcGIS for temperature data (esp. Krigging)
• Further research on how to measure vulnerability (esp. measuring at
ecoregion vs. ecosystem scale for Panama)
32
1 & 2 February (8:30-16:30)
• Further literature review
o Included 2001 IPCC Report (Vol. 2: Impacts, Adaptation, and
Vulnerability)
• Meeting with Dan Irwin to explain project ideas and to learn what high
resolution climate change models are available (1 Feb, 10:00)
• Teleconference with Jean-Nicolas Poussart in Havana (GIS expert in
assessment and management land-based pollution) to discuss how to make
indices
32
Ecosystem Vulnerability to Climate Change 68
68
5 – 9 February (8:30-16:30) Internship Week #1
• Further literature review
• Refined variables to measure ecosystem vulnerability
• Compared UNESCO’s Arc data to its written descriptions on Panama’s
ecosystems
• Informal “Progress Report” presentation to Emil Cherrington: the study
question and in what domains research will be done, goals, objectives,
hypotheses, methods, concerns, and final products (9 Feb, 11:00)
80
12 – 14 February (8:30-16:30) Internship Week #2
• Further literature review
• Preparation of Work Plan – Progress Report (due 14 Feb 18:30)
80
1 & 2 March (8:30-16:30)
• Finalizing research and project development
• SERVIR workshop
32
8 & 9 March (8:30-16:30)
• Measuring EVCC variables – determining vulnerability to sea level rise
32
15 March (8:30-16:30)
• Measuring EVCC variables – determining vulnerability to sea level rise
• Preparing informal presentation & mock-presentation to Emil Cherrington
16
16 March
• Informal presentation of projects to other PFSS internship groups
0
17 March (13:00-17:00)
• Measuring EVCC variables – summarizing sea level rise vulnerability &
preparing climatic ‘space’ analysis
4
19 – 22 March (8:30-16:30) Internship Week #3
• Measuring EVCC variables – summarizing sea level rise vulnerability,
analyzing ecosystem geometries, & analyzing climatic ‘space’
• Experts workshop – sat in on an meeting for monitoring climate change
and SERVIR information sessions; described project to Dr. Carey Yeager
of USAID
64
25 March (13:00-17:00)
• Measuring EVCC variables – analyzing ecosystem geometries
4
29 & 30 March (8:30-16:30)
• Measuring EVCC variables – analyzing ecosystem geometries & climatic
‘space’
32
2 & 3 April (13:00-17:00)
• Measuring EVCC variables – analyzing ecosystem geometries & climatic
‘space’
16
4 April (13:00-19:00)
• Measuring EVCC variables – analyzing ecosystem geometries & climatic
‘space’
• Information meeting on symposium and final projects, brief meeting with
Dr. Roberto Ibáñez to discuss presentation
12
Ecosystem Vulnerability to Climate Change 69
69
5 & 6 April (9:00-18:00)
• Measuring EVCC variables – analyzing climatic ‘space’ & preparing
species sensitivity analysis
• Measurement of degree of human intervention for ecosystems
• Overlaying EVCC variables with protected areas and degree of human
intervention
36
7 April (10:00-18:00)
• Measuring EVCC variables – analyzing climatic ‘space’ & species
sensitivity
• Overlaying EVCC variables with protected areas and degree of human
intervention
16
16 – 20 April (8:30-16:30) Internship Week #4
• Creating final products (maps, final report)
80
21 & 22 April (10:00-18:00)
• Creating final products
32
23 - 25 April (8:30-17:30)
• Completion of final products
• Presentation of results to CATHALAC
54
• 0
26 April
• Submission of final products
Internship symposium (8:00)
Days spent (not necessarily full): 50 Total person-hours: 721
Note: Most, but not all, days involved both researchers working simultaneously. This is
reflected in the person-hours.
Ecosystem Vulnerability to Climate Change 70
70
Appendix 5. Notes of Gratitude – Contact Information
CATHALAC: Attn: Emil Cherrington, Roxana Segundo
Tel: (507) 317-3200
Fax: (507) 317-3299
E-mail: [email protected], [email protected]
Address: Ciudad del Saber, Edificio 801, Clayton, Panama
Website: http://www.cathalac.org
NASA: Attn: Daniel Irwin
E-mail: [email protected]
Address: Marshall Space Flight Center, Building 4200, Room 120, MSFC,
Huntsville, Alabama 35812
Website: http://www.nasa.gov